WO2021041931A1 - Methods of producing dna methylation profiles - Google Patents

Abstract

A method for producing a DNA methylation profile of a subject having or suspected of having systemic lupus erythematosus (SLE) is provided, where DNA methylation is assessed for a gene or panel of genes comprising one or more (e.g., two or more) of PDE7A, PARP14, IFI44L, MX1, PARP9, EPSTI1, RSAD2, IFI27 and B2M. Also provided are methods of administering to a subject identified as having a DNA methylation profile characteristic of an SLE severity subgroup, as defined herein, a therapy indicated for the SLE severity subgroup. Also provided are methods of treating a subject having a disorder characterized by type I interferon signaling, the methods comprising administering a therapeutic amount of an anti-type I interferon signaling therapy to a subject identified as having a DNA methylation profile characteristic of a disorder characterized by type I interferon signaling.

Description

METHODS OF PRODUCING DNA METHYLATION PROFILES

STATEMENT OF GOVERNMENT SUPPORT

This invention was made with Government support under contract number U01 DP005120 awarded by the Centers for Disease Control and Prevention, and contract numbers P30AR070155 and P60AR053308 awarded by The National Institutes of Health. The Government has certain rights in the invention.

CROSS REFERENCE TO RELATED APPLICATIONS

This application claims the benefit of priority under 35 U.S.C. §119(e) of U.S. provisional patent application serial no. 62/893,019 filed on August 28, 2019, the disclosure and attachments thereof are hereby incorporated by reference in their entirety.

INTRODUCTION

Systemic lupus erythematosus (SLE) is an autoimmune disease which is caused by the 'self-attack' by the immune system against the body and results in inflammation and tissue damaged. SLE can manifest in a chronic manner or be of a form that has recurrent relapses. SLE is considered to be a prototypic systemic autoimmune disease - it has the potential of affecting multiple organ systems including the skin, muscles, bones, lungs, kidneys, cardiovascular and central nervous systems. Renal complications, infections, myocardial infarction and central nervous system involvement are the major causes of morbidity and even death in SLE patients. The extremely diverse and variable clinical manifestations present a challenge on the SLE management to clinicians. In addition, outcomes of SLE vary among different racial groups.

The present invention provides means for identifying SLE disease status of a SLE patients irrespective of the ethnicity of the SLE patient.

SUMMARY

Provided are methods of producing a DNA methylation profile of a subject. The methods include assessing DNA methylation in a sample obtained from a subject having or suspected of having systemic lupus erythematosus (SLE), where DNA methylation is assessed for a gene or panel of genes comprising one or more (e.g., two or more) of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M. Also provided are methods of treating a subject having systemic lupus erythematosus (SLE). Such methods comprise administering to a subject identified as having a DNA methylation profile characteristic of an SLE severity subgroup, as defined herein, a therapy indicated for the SLE severity subgroup, where the DNA methylation profile comprises the DNA methylation status for a gene or panel of genes comprising one or more (e.g., two or more) of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M. Also provided are methods of treating a subject having a disorder characterized by type I interferon signaling, the methods comprising administering a therapeutic amount of an anti-type I interferon signaling therapy to a subject identified as having a DNA methylation profile characteristic of a disorder characterized by type I interferon signaling.

Other embodiments are directed to computers, systems and computer readable media associated with methods described herein.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 : Integrative Analysis Pipeline. An overview of the omics data integration strategy used to characterize clinical clusters identified by K-means clustering. MCA = Multiple Component Analysis, FIWE= Flardy-Weinberg Equilibrium, MAF= minor allele frequency, LD= linkage disequilibrium, FDR= false discovery rate, meQTL= cis-methylation quantitative trait loci.

FIGS. 2A-2C: Characterization of clinical features between clusters. A. Distribution of American College of rheumatology (ACR) classification criteria for SLE within each cluster where red indicates presence and blue absence of each criterion. Association between each criterion and cluster was evaluated by a Fisher exact test. B. Criteria significantly associated with each cluster (FDR <0.01). C. Distribution of lupus severity index across clusters with p-value computed using an ANOVA test.

FIGS. 3A-3B: Cluster associated CpGs and meQTL associations. A. Heatmap of CpGs significantly associated with clinical cluster (FDR<0.1) B. Manhattan plot shows - logio(p-value) for associations between cluster-associated CpGs and all SNPs within 1 Mb of each CpG. For each CpG with a significant meQTL (FDR<0.05), the most significant variant is labelled with its corresponding gene.

FIG. 4. Validation of differentially methylated CpGs between clusters Differentially methylated CpGs between clusters were validated in an external dataset (19) A. cluster S1 vs M. B. cluster S2 vs S1 C. cluster S2 vs M. Difference in CpG beta values for differentially methylated sites in CLUES data on x axis and corresponding delta beta value for CpG in validation data on y axis. Green color indicates significant association with cluster in validation dataset (FDR<0.1).

FIG. 5. Identification of candidate cluster-associated cg07259759 ( USP35) that mediates genetic association of rs7104222 ( GAB2 ) with clusters. Association of DNA methylation of cg07259759 and cluster (A) or genotype of rs7104222 (B). C. Association between genotype rs7104222 and clusters. D. Beta coefficient represents the dependence of cluster on genotype with or without adjusting for methylation. Error bars represent the 95% confidence interval for beta coefficient estimate. After adjusting for methylation, the observed dependence reduces towards zero.

FIG. 6. Enrichment of ethnicity-associated CpGs in set of cluster-associated CpGs. A. Null distribution generated by randomly permuting ethnicity labels 1000 times and identifying the number of cluster associated CpGs that were also significantly with ethnicity (p <0.05) in each sample. The red line indicates the number of significant ethnic-associated CpGs (238) found in the set of 256 cluster-associated CpGs. B. Working model illustrating the role of ethnic-associated non-genetic factors in controlling both SLE disease subtypes and methylation signature.

DETAILED DESCRIPTION

Before the methods of the present disclosure are described in greater detail, it is to be understood that the methods are not limited to particular embodiments described, as such may, of course, vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting, since the scope of the methods will be limited only by the appended claims.

Where a range of values is provided, it is understood that each intervening value, to the tenth of the unit of the lower limit unless the context clearly dictates otherwise, between the upper and lower limit of that range and any other stated or intervening value in that stated range, is encompassed within the methods. The upper and lower limits of these smaller ranges may independently be included in the smaller ranges and are also encompassed within the methods, subject to any specifically excluded limit in the stated range. Where the stated range includes one or both of the limits, ranges excluding either or both of those included limits are also included in the methods.

Certain ranges are presented herein with numerical values being preceded by the term “about.” The term “about” is used herein to provide literal support for the exact number that it precedes, as well as a number that is near to or approximately the number that the term precedes. In determining whether a number is near to or approximately a specifically recited number, the near or approximating unrecited number may be a number which, in the context in which it is presented, provides the substantial equivalent of the specifically recited number.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the methods belong. Although any methods similar or equivalent to those described herein can also be used in the practice or testing of the methods, representative illustrative methods are now described.

All publications and patents cited in this specification are herein incorporated by reference as if each individual publication or patent were specifically and individually indicated to be incorporated by reference and are incorporated herein by reference to disclose and describe the materials and/or methods in connection with which the publications are cited. The citation of any publication is for its disclosure prior to the filing date and should not be construed as an admission that the present methods are not entitled to antedate such publication, as the date of publication provided may be different from the actual publication date which may need to be independently confirmed.

It is noted that, as used herein and in the appended claims, the singular forms “a”, “an”, and “the” include plural referents unless the context clearly dictates otherwise. It is further noted that the claims may be drafted to exclude any optional element. As such, this statement is intended to serve as antecedent basis for use of such exclusive terminology as “solely,” “only” and the like in connection with the recitation of claim elements, or use of a “negative” limitation.

It is appreciated that certain features of the methods, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the methods, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable sub-combination. All combinations of the embodiments are specifically embraced by the present disclosure and are disclosed herein just as if each and every combination was individually and explicitly disclosed, to the extent that such combinations embrace operable processes and/or compositions. In addition, all sub-combinations listed in the embodiments describing such variables are also specifically embraced by the present methods and are disclosed herein just as if each and every such sub-combination was individually and explicitly disclosed herein.

As will be apparent to those of skill in the art upon reading this disclosure, each of the individual embodiments described and illustrated herein has discrete components and features which may be readily separated from or combined with the features of any of the other several embodiments without departing from the scope or spirit of the present methods. Any recited method can be carried out in the order of events recited or in any other order that is logically possible. DEFINITIONS

A "site" corresponds to a single site, which may be a single base position or a group of correlated base positions, e.g., a CpG site. A "locus" may correspond to a region that includes multiple sites. A locus can include just one site, which would make the locus equivalent to a site in that context.

A "methylation profile" (also referred to as methylation status) includes information related to DNA methylation for a region. Information related to DNA methylation can include, but is not limited to, a methylation index of a CpG site, a methylation density of CpG sites in a region, a distribution of CpG sites over a contiguous region, and a pattern or level of methylation for each individual CpG site within a region that contains more than one CpG site, and non-CpG methylation. The latter can involve the methylation of cytosine that precede a base other than G, including A, C or T. A methylation profile of a substantial part of the genome can be considered equivalent to the methylome. "DNA methylation" in mammalian genomes typically refers to the addition of a methyl group to the 5' carbon of cytosine residues (i.e. 5-methylcytosines) among CpG dinucleotides. DNA methylation may occur in cytosines in other contexts, for example CHG and CHH, where H is adenine, cytosine or thymine. Cytosine methylation may also be in the form of 5 - hydroxymethylcytosine. Non-cytosine methylation, such as N6-methyladenine, has also been reported.

The "methylation index" for each genomic site (e.g., a CpG site) refers to the proportion of sequence reads showing methylation at the site over the total number of reads covering that site. The "methylation density" of a region is the number of reads at sites within the region showing methylation divided by the total number of reads covering the sites in the region. The sites may have specific characteristics, e.g., being CpG sites. Thus, the "CpG methylation density" of a region is the number of reads showing CpG methylation divided by the total number of reads covering CpG sites in the region (e.g., a particular CpG site, CpG sites within a CpG island, or a larger region). For example, the methylation density for each 100-kb bin in the human genome can be determined from the total number of cytosines not converted after bisulfite treatment (which corresponds to methylated cytosine) at CpG sites as a proportion of all CpG sites covered by sequence reads mapped to the 100-kb region. This analysis can also be performed for other bin sizes, e.g. 50-kb or 1-Mb, etc. A region could be the entire genome or a chromosome or part of a chromosome (e.g. a chromosomal arm). The methylation index of a CpG site is the same as the methylation density for a region when the region only includes that CpG site. The "proportion of methylated cytosines" refers the number of cytosine sites, "C's", that are shown to be methylated (for example unconverted after bisulfite conversion) over the total number of analyzed cytosine residues, i.e. including cytosines outside of the CpG context, in the region. The methylation index, methylation density and proportion of methylated cytosines are examples of "methylation levels."

A "biological sample" refers to any sample that is taken from a subject (e.g., a human, such as a person with SLE, or a person suspected of having SLE) and contains one or more nucleic acid molecule(s) of interest. The biological sample can be a bodily fluid, such as blood, plasma, serum, urine, vaginal fluid, uterine or vaginal flushing fluids, pleural fluid, ascitic fluid, cerebrospinal fluid, saliva, sweat, tears, sputum, bronchoalveolar lavage fluid, etc. Stool samples can also be used. In certain embodiments, the biological sample includes peripheral blood mononuclear cell (PBMC) sample.

A “gene,” for the purposes of the present disclosure, includes a DNA region encoding a gene product, as well as all DNA regions which regulate the production of the gene product, whether or not such regulatory sequences are adjacent to coding and/or transcribed sequences. Accordingly, a gene includes, but is not necessarily limited to, promoter sequences, terminators, translational regulatory sequences such as ribosome binding sites and internal ribosome entry sites, enhancers, silencers, insulators, boundary elements, replication origins, matrix attachment sites and locus control region.

The terms “patient” or “subject” are used interchangeably to refer to a human or a non-human animal (e.g., a mammal).

The terms “treat”, “treating”, treatment” and the like refer to a course of action initiated after a disease, disorder or condition, or a symptom thereof, has been diagnosed, observed, and the like so as to eliminate, reduce, suppress, mitigate, or ameliorate, either temporarily or permanently, at least one of the underlying causes of a disease, disorder, or condition afflicting a subject, or at least one of the symptoms associated with a disease, disorder, condition afflicting a subject.

METHODS OF PRODUCING A DNA METHYLATION PROFILE OF A SUBJECT

Systemic lupus erythematous (SLE) is a heterogeneous autoimmune disease in which outcomes vary among different racial groups. Disclosed herein are methylation profiles of SLE patients divided into three patient clusters that varied according to disease severity. Methylation association analysis identified a set of 256 differentially methylated CpGs across these three patient clusters, including 101 CpGs in genes in the Type I Interferon pathway. Thus, DNA methylation in these sites constitute a biomarker for disease severity and therapeutic intervention.

In certain embodiments, the of producing a DNA methylation profile of a subject may include assessing DNA methylation in a sample obtained from a subject having or suspected of having systemic lupus erythematosus (SLE), wherein DNA methylation is assessed for a gene or panel of genes comprising one or more of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M, to produce a DNA methylation profile of the subject. In certain aspects, the method may also include assessing DNA methylation for other genes which genes one or more of the other genes listed in Fig. 3B.

In certain embodiments, DNA methylation is assessed for a panel of genes comprising two or more, three or more, four or more, five or more, six or more, seven or more, eight or more, or each of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

In certain embodiments, the method further comprises assigning the subject to an SLE severity subgroup based on the DNA methylation profile. The SLE severity subgroup may be “mild” or “M” or severe “S”. In certain embodiments, the S SLE severity subgroup may be severe 1 or “S1” or severe 2 or “S2”. As explained in the examples section, SLE patients in SLE severity subgroup M generally show malar rash, oral ulcers, and late-onset of SLE; SLE patients in SLE severity subgroup S1 have detectable anti-dsDNA and anti- Sm antibodies in blood, lupus nephritis, and early-onset of SLE; and SLE patients in SLE severity subgroup S2 have detectable anti-dsDNA and anti-Sm antibodies in blood, lupus nephritis, early-onset of SLE, leukopenia, lymphopenia, thrombocytopenia, and myocarditis.

In certain embodiments, the method further comprises assigning the SLE severity subgroup “mild” to the subject based on the DNA methylation profile, where the DNA methylation profile comprises hypermethylation of one or more the genes IFI44L, MX1, PARP9, EPSTI1, or PDE7A. As used herein, hypermethylation refers to presence of an increased methylation level (e.g., average methylation values and/or number of methylated CpG sites) in a gene as compared to the same in the same gene in a subject known to have a more severe SLE severity subgroup, such as, severity 1 “S1” or severity 2 “S2”. Presence or absence of hypermethylation can be determined based upon a comparison of methylation level of a gene in the subject to a reference methylation level for the gene, which reference methylation level includes (i) methylation level for hypermethylation of the gene (e.g., obtained from a subject known to have mild SLE severity) and/or (ii) methylation level for hypomethylation of the gene (e.g., obtained from a subject known to have S1 or S2 SLE severity).

In certain embodiments, the method further comprises assigning the SLE severity subgroup “severe” to the subject based on the DNA methylation profile, where the DNA methylation profile comprises hypomethylation of one or more the genes IFI44L, MX1, PARP9, EPSTI1, or PDE7A. As used herein, hypomethylation refers to presence of a reduced methylation level (e.g., average methylation values and/or number of methylated CpG sites) in a gene as compared to the same in the same gene in a subject known to have a less severe SLE severity subgroup, such as, mild “M”. Presence or absence of hypomethylation can be determined based upon a comparison of methylation level of a gene to a reference methylation level for the gene, which reference methylation level includes (i) methylation level for hypermethylation of the gene (e.g., obtained from a subject known to have mild SLE severity) and/or (ii) methylation level for hypomethylation of the gene (e.g., obtained from a subject known to have S1 or S2 SLE severity).

Differentially methylated CpGs in IFI44L, MX1, and PARP9 may be present in the 5- UTR region of the genes. Differentially methylated CpGs in EPSTI1 and PDE7A may be located in the gene body.

In certain embodiments, the panel of genes comprises 500 or fewer genes, such as, 250 or fewer, 150 or fewer, 100 or fewer, 75 or fewer, or 50 or fewer genes.

The methods find use in monitoring treatment or progression of SLE in a patient. In certain embodiments, a method for monitoring efficacy of a treatment for SLE in a patient is provided. The method may include generating a DNA methylation profile of a SLE patient using a first sample obtained prior to start of a treatment for SLE and generating a DNA methylation profile of the SLE patient using a second sample obtained after the treatment for SLE, where a DNA methylation profile that changes from S to M is indicative of efficacy of the treatment. The second sample may be obtained after the treatment regimen has been ongoing for at least a week, a month, 6 months, a year, or more.

In certain embodiments, a method for monitoring progression of SLE in a patient is provided. The method may include generating a DNA methylation profile of a SLE patient using a sample obtained at a first time point and generating a DNA methylation profile of the SLE patient using a sample obtained at a second time point which occurs after the first time point, where a DNA methylation profile that changes from M to S is indicative of progression of SLE. The second time point may be at least a week, a month, 6 months, a year, or more after the first time point.

In certain embodiments, the method may include a step of obtaining a sample from the subject. In other embodiments, the sample may have been obtained prior to the assaying, e.g., at a location remote from the location where assessing is performed. The sample may have been pre-processed to extract genomic DNA by methods known to those of skill in the art. Examples include using the QIAamp® DNA blood Mini Kit or a Qiagen DNeasy Blood extraction kit (both commercially available from Qiagen, Gaithersburg, Md., USA) to extract genomic DNA.

DNA methylation can be assessed by using a range of methylation-aware platforms, including but not limited to MethylationEPIC BeadChip (from lllumina), massively parallel sequencing, single molecular sequencing, bisulphite sequencing, microarray (e.g. oligonucleotide arrays), nanopore-based DNA sequencing system, or mass spectrometry (such as the Epityper, Sequenom, Inc., analysis). In some embodiments, such analyses may be preceded by procedures that are sensitive to the methylation status of DNA molecules, including, but not limited to, cytosine immunoprecipitation and methylation- aware restriction enzyme digestion.

TREATMENT OF SLE

Treatment of SLE include various immunosuppressive agents such as cyclophosphamide, methotrexate, and mycophenolate mofetil. Treatments for mild SLE include nonsteroidal anti-inflammatory drugs (NSAID) and analgesics for fever, arthralgia and arthritis, and topical sunscreens to minimize photosensitivity. Examples of NSAID include acetylsalicylic acid (e.g., aspirin), ibuprofen (Motrin), naprosyn, indomethacin (Indocin), nabumetone (Relafen), and tolmetin (Tolectin). Examples of analgesics include acetaminophen (e.g., Tylenol).

Additional treatments include antimalarials (such as hydroxychloroquine (Plaquenil)) and corticosteroids (such as prednisone) to control joint pain, arthritis, and rash. Patients with moderate to severe disease activity are treated with antimalarials or corticosteroids and steroid dose-reducing agents such as azathioprine, cyclophosphamide, or mycophenolate mofetil. Further examples of antimalarials include Chloroquine (Aralen). Azathioprine (Imuran), mycophenolate mofetil (CellCept), and cyclophosphamide (Cytoxan) are immune suppressors that act in a manner similar to corticosteroids to suppress inflammation and suppress the immune system.

Additional treatments include administering an anti-IFNa antibody. MEDI-545 (alos referred to as M DX-1103) is a fully human 147,000 dalton IgG 1 k monoclonal antibody (Mab) that binds to many interferon a (IFNa) subtypes.

In certain embodiments, a method of treating a subject having systemic lupus erythematosus (SLE) is disclosed. The method may include administering to a subject identified as having a DNA methylation profile characteristic of an SLE severity subgroup a therapy indicated for the SLE severity subgroup, wherein the DNA methylation profile comprises the DNA methylation status for a gene or panel of genes comprising one or more of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

Also provided herein is a method of treating a subject having a disorder characterized by type I interferon signaling. The method comprising administering a therapeutic amount of an anti-type I interferon signaling therapy to a subject identified as having a DNA methylation profile characteristic of a disorder characterized by type I interferon signaling, wherein the DNA methylation profile comprises the DNA methylation status for a gene or panel of genes comprising one or more of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M. A disorder characterized by type I interferon signaling may be SLE, rheumatoid arthritis, inflammatory bowel disease, cancer, or Alzheimer’s disease.

The DNA methylation profile may include the DNA methylation status for a panel of genes comprising two or more, three or more, four or more, five or more, six or more, seven or more, or each of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

In certain embodiments, the DNA methylation profile comprises the DNA methylation status for a panel of genes comprising two or more, three or more, four or more, five or more, six or more, seven or more, eight or more, or each of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

In certain embodiments, the panel of genes comprises 500 or fewer genes, such as, 250 or fewer, 150 or fewer, 100 or fewer, 75 or fewer, or 50 or fewer genes.

In certain embodiments, administering an anti-type I interferon signaling therapy to the subject may include administering an anti-interferon-a agent to the subject, where the anti-interferon-a agent is an antibody. In certain embodiments, the antibody may be sifalimumab or an active fragment thereof.

In certain embodiments, administering an anti-type I interferon signaling therapy to the subject comprises administering an anti-type 1 interferon receptor agent to the subject. The agent may be an antibody, an inhibitory RNA, e.g., shRNA or siRNA and antisense RNA and the like.

As indicated herein, identifying a patient as having mild SLE based on the DNA methylation profile will inform on treatment options for that patient, e.g., a patient identified as having SLE severity subgroup M, may not be treated with corticosteroid and hence can avoid the side effects, such as heart damage, associated with long term corticosteroid treatment. On the other hand, identifying a patient as having severe SLE (SLE severity subgroup S1 or S2) based on the DNA methylation profile will ensure that the patient immediately receives corticosteroid treatment to avoid damage to other organs, such as, kidneys. A patient with severe SLE and receiving corticosteroid treatment can be monitored by DNA methylation profiling and the treatment switched to non-steroidal treatment once the DNA methylation profile indicates that the patient has improved and has a DNA methylation profile indicative of mild SLE.

In certain aspects, a treatment decision may be based upon DNA methylation profile of a SLE patient. For example, if based on presence of hypermethylation of one or more of the genes IFI44L, MX1, PARP9, and EPSTI1, a subject is assigned the SLE severity subgroup mild, this subject may not be treated by administering anti-IFNa antibody. In certain embodiments, a subject is assigned the SLE severity subgroup S (e.g., S1 or S2) based on presence of hypomethylation of one or more of the genes IFI44L, MX1, PARP9, and EPSTI1 and the treatment for this subject may include administering anti-IFNa antibody.

In certain aspects, if based on presence of hypermethylation of one or more of the genes IFI44L, MX1, PARP9, and EPSTI1, a subject is assigned the SLE severity subgroup mild, this subject may be treated by administering NSAID and/or analgesics, wherein the administering may be self-administering.

In certain embodiments, a method of treating a subject assigned the SLE severity subgroup S (e.g., S1 or S2) based on presence of hypomethylation of one or more of the genes IFI44L, MX1, PARP9, and EPSTI1 may include administering antimalarials and/or corticosteroids and steroid dose-reducing agents.

COMPUTER-READABLE MEDIA AND SYSTEMS

According to some embodiments, the methods of the present disclosure are computer-implemented. By "computer-implemented” is meant at least one step of the method is implemented using one or more processors and one or more non-transitory computer-readable media. For example, in certain embodiments, provided are computer- implemented methods for producing a DNA methylation profile of a subject, the methods being implemented using one or more processors and one or more non-transitory computer-readable media comprising instructions stored thereon, which when executed by the one or more processors, cause the one or more processors to assess methylation levels determined from a sample obtained from a subject having or suspected of SLE. The computer-implemented methods of the present disclosure may further comprise one or more steps that are not computer-implemented, e.g., obtaining a sample (e.g., a blood sample) from the subject, preparing the sample for nucleic acid sequencing, determination of methylation level, and/or the like. In certain embodiments, the assessing step may be computer-implemented such that it is performed using one or more processors and one or more non-transitory computer-readable media comprising instructions stored thereon, which when executed by the one or more processors, cause the one or more processors to assess DNA methylation levels. For example, the instructions may cause the one or more processors to compare each of the determined methylated sequences to methylation values stored on a computer-readable medium or a database and determine whether the DNA is hypermethylation or hypomethylated in comparison to the methylation values stored on a computer-readable medium or a database and generate a DNA methylation profile that includes information regarding methylation status of one or more of the genes disclosed herein.

A variety of processor-based systems may be employed to implement the embodiments of the present disclosure. Such systems may include system architecture wherein the components of the system are in electrical communication with each other using a bus. System architecture can include a processing unit (CPU or processor), as well as a cache, that are variously coupled to the system bus. The bus couples various system components including system memory, (e.g., read only memory (ROM) and random access memory (RAM), to the processor.

System architecture can include a cache of high-speed memory connected directly with, in close proximity to, or integrated as part of the processor. System architecture can copy data from the memory and/or the storage device to the cache for quick access by the processor. In this way, the cache can provide a performance boost that avoids processor delays while waiting for data. These and other modules can control or be configured to control the processor to perform various actions. Other system memory may be available for use as well. Memory can include multiple different types of memory with different performance characteristics. Processor can include any general purpose processor and a hardware module or software module, such as first, second and third modules stored in the storage device, configured to control the processor as well as a special-purpose processor where software instructions are incorporated into the actual processor design. The processor may essentially be a completely self-contained computing system, containing multiple cores or processors, a bus, memory controller, cache, etc. A multi-core processor may be symmetric or asymmetric.

To enable user interaction with the computing system architecture, an input device can represent any number of input mechanisms, such as a microphone for speech, a touch- sensitive screen for gesture or graphical input, keyboard, mouse, motion input, speech and so forth. An output device can also be one or more of a number of output mechanisms. In some instances, multimodal systems can enable a user to provide multiple types of input to communicate with the computing system architecture. A communications interface can generally govern and manage the user input and system output. There is no restriction on operating on any particular hardware arrangement and therefore the basic features here may easily be substituted for improved hardware or firmware arrangements as they are developed.

The storage device is typically a non-volatile memory and can be a hard disk or other types of computer-readable media which can store data that are accessible by a computer, such as magnetic cassettes, flash memory cards, solid state memory devices, digital versatile disks, cartridges, random access memories (RAMs), read only memory (ROM), and hybrids thereof.

The storage device can include software modules for controlling the processor. Other hardware or software modules are contemplated. The storage device can be connected to the system bus. In one aspect, a hardware module that performs a particular function can include the software component stored in a computer-readable medium in connection with the necessary hardware components, such as the processor, bus, output device, and so forth, to carry out various functions of the disclosed technology.

Embodiments within the scope of the present disclosure may also include tangible and/or non-transitory computer-readable storage media or devices for carrying or having computer-executable instructions or data structures stored thereon. Such tangible computer-readable storage devices can be any available device that can be accessed by a general purpose or special purpose computer, including the functional design of any special purpose processor as described above. Byway of example, and not limitation, such tangible computer-readable devices can include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other device which can be used to carry or store desired program code in the form of computer- executable instructions, data structures, or processor chip design. When information or instructions are provided via a network or another communications connection (either hardwired, wireless, or combination thereof) to a computer, the computer properly views the connection as a computer-readable medium. Thus, any such connection is properly termed a computer-readable medium. Combinations of the above should also be included within the scope of the computer-readable storage devices.

Computer-executable instructions include, for example, instructions and data which cause a general purpose computer, special purpose computer, or special purpose processing device to perform a certain function or group of functions. Computer-executable instructions also include program modules that are executed by computers in stand-alone or network environments. Generally, program modules include routines, programs, components, data structures, objects, and the functions inherent in the design of special- purpose processors, etc. that perform tasks or implement abstract data types. Computer- executable instructions, associated data structures, and program modules represent examples of the program code means for executing steps of the methods disclosed herein. The particular sequence of such executable instructions or associated data structures represents examples of corresponding acts for implementing the functions described in such steps. Other embodiments of the disclosure may be practiced in network computing environments with many types of computer system configurations, including personal computers, hand-held devices, multi-processor systems, microprocessor-based or programmable consumer electronics, network PCs, minicomputers, mainframe computers, and the like. Embodiments may also be practiced in distributed computing environments where tasks are performed by local and remote processing devices that are linked (either by hardwired links, wireless links, or by a combination thereof) through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.

Notwithstanding the appended claims, the present disclosure is also defined by the following embodiments:

1 . A method of producing a DNA methylation profile of a subject, comprising: assessing DNA methylation in a sample obtained from a subject having or suspected of having systemic lupus erythematosus (SLE), wherein DNA methylation is assessed for a gene or panel of genes comprising one or more of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M, to produce a DNA methylation profile of the subject.

2. The method according to embodiment 1 , wherein DNA methylation is assessed for a panel of genes comprising two or more, three or more, four or more, five or more, six or more, seven or more, eight or more, or each of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

3. The method according to embodiment 1 or embodiment 2, further comprising assigning the subject to an SLE severity subgroup based on the DNA methylation profile.

4. The method according to embodiment 3, further comprising administering to the subject an SLE therapy indicated for the SLE severity subgroup.

5. The method according to any one of embodiments 1 to 4, wherein the panel of genes comprises 500 or fewer genes.

6. The method according to any one of embodiments 1 to 4, wherein the panel of genes comprises 250 or fewer, 150 or fewer, 100 or fewer, 75 or fewer, or 50 or fewer genes. 7. The method according to any one of embodiments 1 to 6, wherein the sample obtained from the subject is a peripheral blood mononuclear cell (PBMC) sample.

8. The method according to any one of embodiments 1 to 7, further comprising obtaining the sample from the subject.

9. The method according to any one of embodiments 1 to 8, wherein the assessing is by bisulphite sequencing.

10. The method according to any one of embodiments 1 to 8, wherein the assessing is by genotyping CpG sites using a microarray.

11. The method according to any one of embodiments 1 to 8, wherein the assessing is by genotyping CpG sites using a DNA sequencing system.

12. The method according to embodiment 11 , wherein the DNA sequencing system is a nanopore-based DNA sequencing system.

13. A method of treating a subject having systemic lupus erythematosus (SLE), comprising: administering to a subject identified as having a DNA methylation profile characteristic of an SLE severity subgroup a therapy indicated for the SLE severity subgroup, wherein the DNA methylation profile comprises the DNA methylation status for a gene or panel of genes comprising one or more of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

14. The method according to embodiment 13, wherein the DNA methylation profile comprises the DNA methylation status for a panel of genes comprising two or more, three or more, four or more, five or more, six or more, seven or more, eight or more, or each of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

15. The method according to embodiment 13 or embodiment 14, wherein the panel of genes comprises 500 or fewer genes. 16. The method according to embodiment 13 or embodiment 14, wherein the panel of genes comprises 250 or fewer, 150 or fewer, 100 or fewer, 75 or fewer, or 50 or fewer genes.

17. A method of treating a subject having a disorder characterized by type I interferon signaling, comprising: administering a therapeutic amount of an anti-type I interferon signaling therapy to a subject identified as having a DNA methylation profile characteristic of a disorder characterized by type I interferon signaling, wherein the DNA methylation profile comprises the DNA methylation status for a gene or panel of genes comprising one or more of IFI44L, MX1 , PARP9, PARP14, EPSTI1 , RSAD2, IFI27 and B2M.

18. The method according to embodiment 17, wherein the DNA methylation profile comprises the DNA methylation status for a panel of genes comprising two or more, three or more, four or more, five or more, six or more, seven or more, or each of IFI44L, MX1 , PARP9, PARP14, EPSTI1 , RSAD2, IFI27 and B2M.

19. The method according to embodiment 17 or embodiment 18, wherein the panel of genes comprises 500 or fewer genes.

20. The method according to embodiment 17 or embodiment 18, wherein the panel of genes comprises 250 or fewer, 150 or fewer, 100 or fewer, 75 or fewer, or 50 or fewer genes.

21. The method according to any one of embodiments 17 to 20, wherein administering an anti-type I interferon signaling therapy to the subject comprises administering an anti- interferon-a agent to the subject.

22. The method according to embodiment 21 , wherein the anti-interferon-a agent is an antibody.

23. The method according to embodiment 22, wherein the antibody is sifalimumab. 24. The method according to any one of embodiments 17 to 20, wherein administering an anti-type I interferon signaling therapy to the subject comprises administering an anti type 1 interferon receptor agent to the subject.

25. The method according to embodiment 24, wherein the anti-type 1 interferon receptor agent is an antibody.

26. The method according to embodiment 25, wherein the antibody is anifrolumab.

The following examples are offered by way of illustration and not by way of limitation.

EXPERIMENTAL

Example 1 - A phenotypic and genomics approach in a multi-ethnic cohort to subtype systemic lupus erythematosus (SLE)

Summary: Systemic lupus erythematous (SLE) is a heterogeneous autoimmune disease in which outcomes vary among different racial groups. Here, we aim to identify SLE subgroups within a multiethnic cohort using an unsupervised clustering approach based on the American College of Rheumatology (ACR) classification criteria. We identify three patient clusters that vary according to disease severity. Methylation association analysis identifies a set of 256 differentially methylated CpGs across clusters, including 101 CpGs in genes in the Type I Interferon pathway, and we validate these associations in an external cohort. A cis-methylation quantitative trait loci analysis identifies 744 significant CpG-SNP pairs. The methylation signature is enriched for ethnic-associated CpGs suggesting that genetic and non-genetic factors may drive outcomes and ethnic-associated methylation differences. Our computational approach highlights molecular differences associated with clusters rather than single outcome measures. This work demonstrates the utility of applying integrative methods to address clinical heterogeneity in multifactorial multi-ethnic disease settings.

Introduction: Systemic lupus erythematosus (SLE) is a multifactorial autoimmune disease with heterogeneous manifestations that encompasses a wide range of disease severity. Complex diseases such as SLE involve a dynamic interplay between molecular processes, many of which are unknown. Long-term outcomes for individual patients are therefore difficult to predict, as is the scope of organ system involvement. While some patients have aggressive disease progression, others do not accrue significant damage within 5 years of SLE diagnosis¹ _’ ² _’ ³ _’ ⁴. We know little about why an affected individual might develop a particular SLE phenotype. Furthermore, a patient can be classified as having SLE if she or he fulfills any four of the 11 American College of Rheumatology (ACR) classification criteria⁵, with resultant extensive disease heterogeneity. In recent years, significant effort has been applied to better sub-classify SLE, not only to predict disease outcomes but also inform specific mechanistic pathways that could be strategically targeted according to subtype^{6 7}’⁸.

SLE disease progression and outcomes vary significantly among different racial/ethnic groups⁹’^{10 11}. Patients from non-European populations, such as Hispanics, African Americans, and Asians, develop SLE at a younger age and experience worse disease manifestations than patients of European descent. Even after decades of basic research and public health initiatives these health disparities remain relatively unchanged. Factors that underlie these disparities are elusive and likely derive in part from complex interactions between genetic and environmental factors, which may in part originate from social inequities. However, the majority of molecular studies to date have been carried out in European populations.

These is evidence that both genetics and DNA methylation play a role in SLE outcomes. Lupus nephritis, a severe outcome of lupus that drives disease mortality, was found to be significantly correlated with genetic variants in ITGAM, TNFSF4, APOL1, PDGFRA, and SLC5A11, among others. The HLA-DR2 and HLA-DR3 alleles have also been associated with susceptibility and autoantibody production in lupus¹² _’ ^{1314 15}. Overall, hypomethylation of interferon-responsive genes has been associated with higher disease activity and renal disease, as well as production of autoantibodies¹⁶’^{17 18}. For example, differentially methylated CpGs in TNK2, DUSP5, MAN1C1, PLEKHA 1, IRF7, HIF3A, IFI44, and PRR4 have been associated with lupus nephritis¹⁸’¹⁹²⁰. Differentially methylated CpGs in I FIT 1, IFI44L, MX1, RSAD2, OAS1, EIF2AK2, PARP9/DTX3L, and RABGAP1L have been associated with production of autoantibodies¹⁶’^{21 22}. However, these studies have been performed largely in patients of European descent.

While numerous previous studies focused on either the genetics or epigenetics of SLE, a multi-omics approach coupled with deep clinical phenotyping may better elucidate the molecular basis of disease heterogeneity. By integrating different layers of molecular and clinical data, several studies have provided insight into mechanisms of complex disease such as Alzheimer’s disease²³’²⁴²⁵, inflammatory bowel disease²⁶, cancer²⁷²⁸, and rheumatoid arthritis²⁹’³⁰’^{31 32}. In this work, we initially apply unsupervised clustering of ACR classification criteria for SLE to define disease subtypes among a diverse multi-ethnic cohort of SLE patients. We then develop and apply an integrative approach leveraging human genetics and DNA methylation data to elucidate differences between these disease subtypes. We find 256 differentially methylated CpGs that varied significantly according to subtype, of which 61 were under proximal genetic control (Fig. 1).

Results

Clinical clustering identifies distinct subtypes of SLE. Clinical characteristics of the 333 patients examined from the UCSF California Lupus Epidemiology Study (CLUES) cohort are presented in Supplementary Table 1 in Lanata, C.M., et al. Nat Commun 10, 3902 (2019). We first stratified SLE patients into clusters based on ACR classification criteria and sub criteria using an unsupervised clustering approach. Briefly, we first applied multiple correspondence analysis (MCA) and then performed K-means clustering on the top two components chosen by a bootstrap resampling strategy (see Methods). Three clusters were identified. The clusters are labelled M (mild), S1 (severe 1) and S2 (severe 2; Fig. 2A, 2B). Cluster M was comprised of 101 patients (30.3%) and was characterized by a high prevalence of malar rash, photosensitivity, arthritis, and serositis, but lower prevalence of hematologic manifestations, lupus nephritis, and serologic manifestations (p< 0.001). Cluster S1 was comprised of 154 patients (46.2%) and was characterized by higher prevalence of lupus nephritis and anti-dsDNA autoantibody positivity (p< 0.001). Cluster S2 was comprised of 78 patients (28.8%) and was the most severe subtype, with a high prevalence of lupus nephritis, autoantibody production (anti- dsDNA, anti-Sm, anti-RNP and antiphospholipid antibodies), and internal organ manifestations, such as hematologic manifestations (Fisher exact test p< 0.001 ; Table 1). The Lupus Severity Index³³, a validated scoring system based on the ACR classification criteria, was also significantly different between the three clusters (ANOVA test p= 2 x 10^~21), with cluster M the least severe, and cluster S2 the most severe (Fig. 2C).

With respect to ethnicity, we found a significant increase in the proportion of White patients in cluster M compared to clusters S1 and S2 (Kruskall— Wallis p= 4.76 c 10^~4), and a higher proportion of Asian patients in clusters S1 and S2 compared to cluster M (Kruskall— Wallis p= 1.4 x 10 ³; Table 1).

Table 1. Summary of significant clinical and demographic variables across clusters. False Discovery Rate (FDR) p-values were calculated for Kruskall-Wallis (continuous variables) or Fisher’s exact test (binary variables).

*ACR= American college of Rheumatology, APLA= antiphospholipid antibodies, FDR= false discovery rate, SLEDAI=SLE disease activity index At the time of blood sampling, patients in the more severe clusters (S1 and S2) had lower levels of complement C3 (ANOVA p= 1.47 10^-3), were more likely to be RNP positive (Fisher exact test p= 3.66 10^~5) and were more likely to be receiving mycophenolate (Fisher exact test p= 2.48 10^~3) and prednisone (Fisher exact test p= 5.07 10^~2) than patients in cluster M (Supplementary Table 2 of Lanata, C.M., et al. Nat Commun 10, 3902 (2019)). We also examined complete blood counts and proportions taken at time of blood draw from all patients and found a statistically significant decrease in leukocytes (ANOVA p = 3.31 c 10^~3), eosinophils (ANOVA p= 3.39 c 10^~2) and lymphocytes (ANOVA p= 2.84 10 ²) among the three clusters (Supplementary Table 2 of

Lanata, C.M., et al. Nat Commun 10, 3902 (2019)). This could represent a marker of disease severity or a consequence of higher immunosuppressant drug use at the time of blood draw for patients in the more severe disease clusters. In a comparison of socioeconomic variables across clinical clusters, we did not observe a statistically significant difference in average education level or income between the three clusters (Supplementary Table 3 of Lanata, C.M., et al. Nat Commun 10, 3902 (2019)).

Distinct methylation patterns distinguish clinical clusters. The clusters identified above, characterized by multiple comorbid phenotypes, represent a clinically relevant framework to stratify SLE patients. Using this stratification, we aimed to identify differentially methylated CpG sites associated with these clinical clusters. Using an ANOVA model, we identified 256 CpG sites in 124 genes that were differentially methylated according to clinical cluster (FDR < 0.1) after adjusting for sex, genetic ancestry principal components, cell composition, medications, alcohol use, and smoking status (Fig. 3A; Supplementary Data 1 of Lanata, C.M., et al. Nat Commun 10, 3902 (2019)). A quantile- quantile plot is shown in Supplementary Figure 1 of Lanata, C.M., et al. Nat Commun 10, 3902 (2019). The observed versus expected test statistic demonstrates no evidence for inflation of the association tests (inflation factor l = 0.99).

Cluster associated CpGs and meQTL associations a Heatmap of CpGs significantly associated with clinical cluster (FDR < 0.1) b Manhattan plot shows -logio(p-value) for associations between cluster-associated CpGs and all SNPs within 1 Mb of each CpG. For each CpG with a significant meQTL (FDR < 0.05), the most significant variant is labelled with its corresponding gene.

Upon mapping these 256 cluster-associated CpG sites to genes and performing pathway analysis, we found significant enrichment of genes associated with Type I interferon signaling, antiviral responses and inflammatory pathways (gene list enrichment analysis FDR < 0.01 ; Table 2).

Table 2. Significantly enriched pathways in cluster-associated CpGs. CpGs were mapped to genes using lllumina annotation file, and pathway analysis was performed using ToppFunn ⁹².

In order to functionally classify the cluster-associated CpGs, we intersected these genomic regions with the Hallmark Interferon-Alpha Responsive gene set³⁴ since the IFN- alpha signaling pathway has been previously implicated in SLE pathology¹⁸’³⁵³⁶’³⁷’³⁸. We observed a significant enrichment of IFN-alpha responsive genes (hypergeometric p< 0.01 ) with 101 out of the 256 CpGs within this set. Notably, of the 101 IFN-alpha CpGs, 93 were hypermethylated in cluster M relative to both cluster S1 and S2. Of these CpGs, 57 were in the promoter region (TSS200, TSS1500, 5' UTR), and 36 were in the gene body. Hypermethylation at the promoter sites suggests a role for epigenetic silencing in cluster M with respect to S1 and S2 while gene body hypermethylation suggests gene expression.

Cluster-associated CpGs with the greatest variance (5-11% methylation variance) across the clusters were in genes IFI44L, MX1, PARP9, EPSTI1, and PDE7A, all displaying hypermethylation in cluster M relative to S1 and S2 (Supplementary Data 2 of Lanata,

C.M., et at. Nat Commun 10, 3902 (2019)). With the exception of PDE7A, all of these genes are interferon responsive. PDE7A encodes a phosphodiesterase associated with T cell activation and IL-2 production³⁹. Differentially methylated CpGs in IFI44L, MX1, and PARP9 map to the 5-UTR region, suggesting silencing of these genes. Differentially methylated CpGs in EPSTI1 and PDE7A are located in the gene body, where hypermethylation is associated with gene expression. For each of the 256 CpGs identified above using the ANOVA test, we then sought to determine which pairwise comparison (cluster S2 vs M, S2 vs S1 , or S1 vs M) contributed to the significant F-statistic. Using the nestedF method in the Limma R package40, 247 of the aforementioned associated 256 CpGs were differentially methylated between clusters S2 and M (FDR p < 0.1 ; Supplementary Fig. 2A, Table 3 and Supplementary Table 5). The most significant CpGs were in the promotor of lFI44L and gene body of RSAD2, with hypermethylation in cluster S1 versus M. Comparison of clusters S2 to S1 identified 18 differentially methylated CpGs (FDR < 0.1 ; Supplementary Fig 2B, Table 3 and Supplementary Table 5), with hypermethylation of CpGs in IFI27 and B2M, a component of the MHC1 complex. Comparison of clusters S1 and M identified 53 differentially methylated CpGs (FDR < 0.1 ; Supplementary Fig. 2C; Table 3 and Supplementary Table 5). The percent variance between clinical clusters explained by CpG methylation varied from 0.9% for cg23002431 {COP A gene body) to 21% for cg00959259 ( PARP 5’UTR).

Table 3. Summary of cluster-wise comparison and validation. Rows indicate individual pairwise comparisons as performed using the nestedF method in Limma.

Validation of cluster-specific methylation profiles. To determine whether the methylation signature associated with the three clinical clusters identified was reproducible, we applied our clustering method to a previously published independent cohort of 302 female SLE patients of European descent¹⁶. This cohort has a lower lupus severity index (6.15 ± 1.42) compared to the CLUES cohort (6.85 ± 1.63; Student’s i test p< 0.001), however this difference is relatively small and not clinically significant. In order to identify methylation associations in this validation cohort, we first assigned a cluster label (M, S1 , or S2) to each patient using the study cohort as a reference. Since the clusters in the study cohort were discovered using an unsupervised approach, we first trained a random forest model on the CLUES data with the ACR features as input. Model parameters were optimized by minimizing the out-of-bag error. The model achieved a minimum out-of-bag error of 12.8%. We then applied this model to the validation cohort of SLE patients of European descent and determined a cluster label for each sample. Clinical characteristics of subjects in each cluster in the validation cohort are presented in Supplementary Table 4 of Lanata, C.M., et al. Nat Commun 10, 3902 (2019). In comparison to the CLUES cohort where the majority of subjects were in cluster S1 , the majority in the validation cohort were in cluster M, reflecting the racial differences between the cohorts.

To determine whether the methylation patterns associated with each cluster were robust and reproducible, we evaluated methylation differences in the independent cohort of 302 female SLE patients of European descent as described above¹⁶. Since the validation dataset was obtained using the lllumina 450k BeadChip, we restricted these analyses to the 158 cluster-associated CpGs in the CLUES cohort that were also on the 450k array. Of these 158 CpGs, 132 (84%) were significantly associated with cluster in the validation dataset (FDR < 0.1 ; Table 3). We observed a strong correlation (r> 0.9) between differences in methylation beta values for the CLUES cohort and validation set for all three pairwise comparisons (cluster S1 vs. M, S2 vs. S1 , and S2 vs. M; Table 3, Fig. 4).

Active chromatin states in cluster-associated CpGs. In order to further characterize the epigenetic landscape of the cluster-associated CpGs, we examined CpG enrichment in genomic regions classified according to specific chromatin states based on the Epigenome Roadmap 15 state model⁴¹. We found significant depletion (Fisher’s exact FDR < 0.01 ; OR < 0.5) of cluster-associated CpGs in quiescent regions in 12 of the 13 peripheral blood cell types, and significant enrichment (FDR < 0.01 ; OR > 2) in enhancers and regions flanking active transcriptional start sites in all cell types. We also observed significant enrichment (Fisher’s exact FDR < 0.01 ; OR > 2) of FI3K4me3, H3K4me1 , and H3K27ac histone marks specific for active enhancers in all peripheral blood cell types.

Epigenetic annotation of differentially methylated CpGs in IFI44L land in enhancers and active transcription sites in peripheral blood primary B cells, T helper memory cells, Naive T cells, Th17 cells, T memory cells and T regs, but not in regulatory or transcription sites in neutrophils or NK cells. Differentially methylated CpGs in MX1, PARP9, EPSTI1, and PDE7A are located in enhancers and transcription sites in most peripheral immune cell subtypes. meQTL loci controlling cluster-associated CpGs. We sought to understand the sources of methylation differences in the clinically-defined clusters. Therefore, we used paired genotype data to investigate genetic drivers of methylation differences. Specifically, we conducted a methylation quantitative trait loci analysis (meQTL) to determine which cluster-associated CpGs were under proximal genetic control (distance between SNP and CpG < 1 Mb). Genetic data was first imputed and LD-pruned (A² < 0.8). After adjusting for population structure, sex, age, cell type composition, medication use, smoking status, and alcohol consumption, we found 744 significant cis meQTL associations (FDR < 0.05; Fig. 3B). These involved 61 unique CpGs in 41 genes, and 397 SNPs in 90 genes.

Of the 744 significant cis meQTL associations, 91 meQTLs in 19 unique CpGs were in interferon-alpha or interferon-gamma responsive genes. Of these, the greatest number of meQTL loci were in EPSTI1 (12 SNPs), PARP14 (nine SNPs), and PARP15 {8 SNPs) for CpGs in interferon-alpha responsive genes. We found 39 meQTL associations involving CpG sites in the promoter region. Notably, we found 21 associations in CpGs in PARP14, of which 20 were in the promoter region under the control of SNPs in PARP14, PARP15, and DIRC2. We also found CpGs in the promoter region of OAS3 (n= 2) under the control of SNPs in LHX5-AS1, and one CpG in USP18 under the genetic control of SNPs in LINC01634.

Of the non-interferon-responsive CpGs, we found 20 genetic variants that controlled methylation of cg07259759 located in the gene body of USP35, a ubiquitin specific peptidase⁴² (methylation variance 21-25%). Ten of these 20 genetic variants were found in an intron of GAB2, a tyrosine kinase adaptor that is primarily upregulated in activated innate immune cells⁴³’⁴⁴’⁴⁵. We also found 43 genetic variants in HLA-F, a MHC-lb minor allele involved in NK cell self-recognition⁴⁶, which controlled methylation at four CpG sites in the gene body of HLA-F. Fifteen CpGs were located in the promoter or 5’UTR region, with the largest methylation variance observed for cg04738877 in the promoter region of GALC, under the control of SNPs in introns of the same gene.

Although we considered all SNPs within a 1 Mb window around each CpG, the proportion of significant meQTL associations decreased as the distance between SNPs and CpGs increased. This suggests that genetically determined CpG methylation was typically driven by proximal genetic variation, rather than distal effects.

Epigenetic mediation of genetic association with clusters. One challenge of interpreting the methylation associations with clusters is that many methylation differences may represent a consequence of clinical differences between clusters rather than causal mediators. In order to identify which CpGs may mediate genetic associations with clusters and reveal novel biology, we employed an integrative causal inference method⁴⁷.

Briefly, this method uses conditional probabilities to evaluate a causal relationship between a factor (genotype), a potential mediator (CpG methylation), and an outcome (clinical cluster).

First, from the list of 744 meQTLs identified, we selected the SNPs that were significantly associated with cluster (FDR < 0.05). For these meQTL associations, we identified the subset where methylation appears to mediate the genotype-cluster association using the causal inference test (CIT). This yielded 24 meQTLs with 21 SNPs (FDR < 0.05). Notably from these, we found 6 significant associations between SNPs in GAB2 and CpGs in USP35. We also found evidence for methylation mediation of SNPs in HLA-F. Figure 5 provides an example of one of these associations between a SNP in GAB2 and CpG in USP35.

Ethnicity-associated differentially methylated CpGs. As some of the methylation differences in the clinically-defined clusters could be explained by genetic variation in the meQTL analysis, we explored the effect of ethnicity, after adjusting for genetic factors. Previous work has identified patterns of differential methylation across ethnic groups due to both ancestral genetic variation and environmental influences⁴⁸. As non-White ethnicity is associated with worse outcomes in SLE, we sought to determine whether the differentially methylated CpGs across clusters were enriched for ethnicity-associated CpGs, after adjusting for genetic ancestry. Of the 256 cluster-associated CpGs, we identified 237 CpGs that were associated with ethnicity (FDR < 0.05) after adjusting for sex, the top three genetic ancestry principal components, cell composition, medications, alcohol use, and smoking status. A permutation analysis was conducted by randomly permuting ethnicity 1000 times and testing for association with ethnicity. Figure 6A displays the density of ethnicity- associated CpGs in 1000 random samples. This analysis revealed a significant (p< 0.001) enrichment of ethnicity-associated CpGs in the cluster-associated methylation signature. Results and data are available as an RShiny Application for use in future research at a web address reachable by typing: http:// followed by comphealth.ucsf.edu/sle_clustering/ in search window of a web browser.

In the present study, we developed a stepwise multi-omics approach for identifying SLE patient subtypes defined by clinically-relevant phenotypes and molecular mechanisms among a multi-ethnic cohort. We report three lupus clinical subtypes defined by the ACR classification criteria that vary according to disease severity. We also show that patterns of differential methylation at specific CpGs are associated with the clinical subtypes. A subset of these CpGs are under genetic control, however the majority display a strong association with ethnicity after adjusting for genetic ancestry, suggesting possible molecular mediators of the ethnic-effect underlying lupus outcomes.

Unlike previous studies that have largely studied SLE patients of European genetic ancestry, we studied a cohort or patients of White, African-American, Asian, and Hispanic ethnicity. Since SLE severity is known to vary widely between racial and ethnic groups, analysis of a large multiethnic cohort is crucial for understanding the genetic and non- genetic determinants of this ethnic-associated variability. Unsupervised clustering approaches have been applied widely to high dimensional omics datasets with the aim of deriving meaningful clusters characterized by a small set of molecular features⁶’^{49 50}. By translating this dimensionality reduction technique to the ACR clinical features in SLE, we identified three clinical subtypes each characterized by specific ACR features. Due to the strong association of DNA methylation with genetic variation, unsupervised clustering of the methylation data revealed population structure rather than lupus-relevant clinical differences. For these reasons, we did not report DNA methylation clustering and rather chose to define subtypes by ACR criteria. Since this is an ongoing cohort, with a larger sample size, we may be able to define methylation-based clusters in each racial group separately, minimizing the effect of genetic structure confounding.

Importantly, the clusters defined in this study are consistent with previous epidemiological studies describing the correlation of multiple sub-phenotypes of SLE, such as the correlation of SLE skin manifestations with arthritis, serositis with the lack of other end-organ involvement, and anti-dsDNA with lupus nephritis^{21 51 52}. The milder subtype in this study had a higher prevalence of participants of White race. This has also been previously described, as patients with European ancestry have a higher proportion of arthritis, skin manifestations and serositis and lower prevalence of lupus nephritis and autoantibody production⁵³⁵⁴’⁵⁵.

By considering clusters defined by multiple phenotypes we preserve the multifactorial clinical nature of SLE. Training a random forest model using the cluster assignments as labels allowed us to apply this clustering scheme to an external dataset of patients of European descent.

After identifying clinically-relevant patient clusters, we found a set of 256 GpGs associated with the clusters with strong enrichment of methylation for genes in the Type I interferon pathway, cytoplasmic viral sensing pathways, and immune related pathways, with significant enrichment in enhancers and regions flanking active transcriptional start sites in all peripheral immune cell types. Several studies have implicated transcriptional upregulation and epigenetic regulation of the Type I interferon pathway in SLE¹⁷’¹⁹’³⁵’³⁶’^{37 56}. We and others have also previously described methylation changes in interferon responsive genes associated with individual lupus outcomes, including cutaneous rash⁵⁷, renal involvement¹⁹’^{20 58} and serologic manifestations^{16 22}. However, given the striking heterogeneity of clinical features in SLE, epigenetic programs associated with single phenotypes may be less relevant in a clinical setting. By performing unsupervised clustering on a diverse SLE cohort, we can study the molecular heterogeneity in clinically-relevant subtypes driven by multiple SLE outcomes and disease severity. With this approach we found that severe subtypes, which also have higher proportions of patients of Asian and Hispanic ethnicity, have a higher degree of type I interferon dysregulation. Of these CpG sites, the greatest methylation variance across the clusters was in IFI44L, which encodes for Interferon Induced Protein 44 Like, with progressive hypomethylation from cluster M to cluster S2. Although the function of IFI44L is unknown, increased IFI44L expression is a component of the type-1 IFN response signature and also part of the cellular response to viral infections⁵⁹. IFI44L promoter methylation has been proposed as a blood biomarker for SLE⁵⁸.

Since genetic variation can have profound effects on DNA methylation⁶⁰⁶¹’⁶², we also performed an meQTL analysis to quantify the degree of proximal genetic control of the 256 CpG signature. Although the cluster-associated CpG set was strongly enriched for Type 1 IFN genes, only a subset of these Type 1 IFN CpGs (24%) had meQTL loci, suggesting that environmental factors may be contributing more to the epigenetic regulation of the Type 1 IFN pathway than genetic factors. We found interesting associations between genetic variation and methylation at CpGs in immune relevant genes, and in 24 meQTL associations we had evidence of mediation of the genetic association by methylation at the corresponding CpG. We would like to highlight variants in HLA-F, PARP14, and GAB2 controlled methylation sites in USP35. HLA-F is part of the nonclassical HLA- Ib genes, which are mono- or oligomorphic⁴⁶. Surface expression of HLA-F has been demonstrated on activated T, B, and NK cells, and serum IgG autoantibodies against HLA- F have been detected in SLE patients and correlated with disease activity⁶³’⁶⁴⁶⁵. PARP14 encodes for poly(ADP-ribose) polymerase (PARP) protein family 14 and assists in posttranslational ribosylation modification of target proteins. Its role in SLE and autoimmune disease has not been defined but it has been shown to regulate glycolysis via IL-4 in B lymphocytes⁶⁶, promote survival of cancer cells⁶⁷, and regulate macrophage activation⁶⁸. Glycolysis in SLE has been found to directly influence the Th17 cell fate and survival, therefore implicating a potential mechanistic role for PARP14 in SLE⁶⁹.

GAB2 is a member of the GRB2-associated binding protein (GAB) gene family. These genes act as adapters for transmitting various signals in response to stimuli through cytokine and growth factor receptors, and T- and B-cell antigen receptors⁴⁵. Among its related pathways is Akt signaling, which is involved in cell proliferation and autophagy, a process that has been implicated in SLE pathogenesis⁴⁴⁷⁰’⁷¹’⁷². Variants of GAB2 influenced methylation marks in the gene body of USP35, which encodes for a member of the peptidase C19 family of ubiquitin-specific proteases⁴². This deubiquitinating enzyme has been shown to mediate the IFN-type I response upon viral infection and it has been associated with higher IFN-b and IFIT1 gene expression⁷³. This is relevant to our findings as higher levels of IFN-b have been associated with SLE⁷⁴ _’ ⁷⁵⁷⁶. Variation in methylation can be attributed to genetic and non-genetic effects. The majority of differentially methylated CpGs among disease subtypes were not classified as under genetic control. Although the number of detected meQTL associations is likely to increase with a larger sample size, it also suggests a greater role for non-genetic or environmental effects. Since self-reported race and ethnicity refers to communality in cultural heritage, language, social practice, traditions, and geopolitical factors, it may be a proxy for a wide variety of environmental exposures not easily captured. Since ethnicity plays a role in differences in lupus severity, we were interested in the association of ethnicity and our methylation findings, after adjusting for genetic variation. After adjusting for genetic principal components, we found a significant enrichment of self-reported race-associated CpGs in our 256 CpG signature (p< 0.001 ). This suggests that there may be a common set of CpGs that mediates both SLE clusters and non-genetic differences in race. Although this might raise concern for confounding by genetic structure, our analysis revealed a low genomic inflation factor. Furthermore, the CpGs are biologically relevant in SLE pathogenesis. In addition, previous work has identified differential methylation between ethnic groups due to environmental factors that is not fully explained by genetic ancestry⁴⁸. It is well known that SLE outcomes vary according to race, and the causes behind these race disparities are a source of ongoing debate. We observed differences in race across the disease subtypes, with the milder subtype having a larger representation of White patients. Therefore, these methylation differences suggests the existence of molecular mediators of the non-genetic race-related clinical differences in SLE outcomes, and may reflect environmental exposures that affect races differentially. Figure 6B displays a model for the role of race-associated non-genetic factors that control both methylation of the 256 CpG signature and SLE disease subtypes.

We applied this approach to an external cohort of exclusively SLE patients of European descent. The clinical clusters were labelled by applying a random forest model trained on the original CLUES clusters. The greatest number of subjects were assigned to cluster M. Since cluster M was also enriched for patients that self-identified as White in the original clusters, we believe that the clustering model accurately captures the role of racial differences in lupus severity. We were only able to test 158 of the 256 cluster-associated CpGs in the validation cohort due to a limited number of overlapping probes. However, we validated 84% of the cluster-associated methylation sites. As we also found that the majority of cluster-associated CpGs were race-associated in the CLUES cohort, this suggests that while race plays a role in DNA methylation, these effects are secondary to the clinical differences between lupus clusters. As we could not determine the relative genetic and non- genetic contributions of the methylation differences, we cannot recreate the race-specific findings we have described for the CLUES cohort. In general, this suggests that the subtypes and associated methylation differences we have described can be applied to other cohorts.

Strengths of this study include the rich phenotyping data and adjustments for major confounders, including medications at the time of blood draw, smoking history, and alcohol consumption, which are unaccounted for in most epigenome-wide association studies. This is also the largest cohort including African American, Caucasian, Asian, and Hispanic patients to be profiled for genome wide DNA methylation and genotyping, which allowed us to differentiate between genetic and non-genetic effects of race in SLE outcomes, shedding light on molecular mediators of race in disease heterogeneity. Future work will include testing these findings in other multi-ethnic cohorts. Furthermore, it will be of interest to determine whether these DNA methylation differences are predictive of future disease activity and severity.

In summary, we have identified three distinct clinical subtypes of SLE that have distinct patterns of methylation at specific CpG sites. While previous studies have defined subtypes based on transcriptomic data⁶⁷⁸, by integrating methylation and genetic data, the three subtypes identified here may reflect the influence of both genetic and non-genetic effects. We also identified potential mediation of genetic association by methylation changes not previously addressed in SLE. Furthermore, we have demonstrated the utility of studying a diverse SLE population to investigate the molecular underpinnings of race differences in SLE outcomes.

Methods

Subjects and samples. Subjects were participants in the California Lupus Epidemiology Study (CLUES), a multi-racial/ethnic cohort of individuals with physician- confirmed SLE. This study was approved by the Institutional Review Board of the University of California, San Francisco. All parficipants signed a written informed consent to participate in the study. Participants were recruited from the California Lupus Surveillance Project, a population-based cohort of individuals with SLE living in San Francisco County from 2007 to 2009²’⁴⁷. Additional participants residing in the geographic region were recruited through local academic and community rheumatology clinics and through existing local research cohorts.

Study procedures involved an in-person research clinic visit, which included collection and review of medical records prior to the visit; a history and physical examination conducted by a physician specializing in lupus; collection of biospecimens, including peripheral blood for clinical and research purposes; and completion of a structured interview administered by an experienced research assistant. All SLE diagnoses were confirmed by study physicians based upon one of the following definitions: (a) meeting > 4 of the 11 American College of Rheumatology (ACR) revised criteria for the classification of SLE as defined in 1982 and updated in 1997⁵·⁷⁷, (b) meeting 3 of the 11 ACR criteria plus a documented rheumatologist’s diagnosis of SLE, or (c) a confirmed diagnosis of lupus nephritis, defined as fulfilling the ACR renal classification criterion (>0.5 grams of proteinuria per day or 3 + protein on urine dipstick analysis) or having evidence of lupus nephritis on kidney biopsy.

DNA methylation assessment. DNA methylation of genomic DNA from peripheral blood mononuclear cells was profiled using the lllumina MethylationEPIC BeadChip. This chip assesses the methylation level of -850,000 CpGs in enhancer regions, gene bodies, promoters, and CpG islands. All subsequent processing was done using the R minfi package. Signal intensities were background subtracted using the minfi noob function and then quantile normalized⁷⁸⁷⁹. Sites with a poor detection rate (detection p value > 0.05) in more than 5% of the samples (1746 CpG sites) were removed. Sites predicted to hybridize to multiple loci (44,097 CpG sites) and sites on non-autosomal chromosomes (19,627 CpG sites) were removed, resulting in 802,579 probes for analyses.

DNA genotyping. Genotyping for genomic DNA from peripheral blood mononuclear cells was performed using the Affymetrix Axiom Genome-Wide LAT 1 Array. This genotyping array is composed of 817,810 SNP markers across the genome and was specifically designed to provide maximal coverage for diverse racial/ethnic populations, including West Africans, Europeans and Native Americans⁸⁰. Samples were retained with Dish QC (DQC) > 0.82. SNP genotypes were first filtered for high-quality cluster differentiation and 95% call rate within batches using SNPolisher. Additional quality control was performed using PLINK. SNPs having an overall call rate less than 95% or discordant calls in duplicate samples were dropped. Samples were dropped for unexpected duplicates in IBD analysis or mismatched sex between genetics and self-report; for first-degree relatives, one sample was retained. All samples had at least 95% genotyping and no evidence of excess heterozygosity (maximum < 2.5*SD). We tested for Hardy-Weinberg Equilibrium (HWE) and cross-batch association for batch effects using a subset of subjects that were of European ancestry and negative for double-stranded-DNA antibodies and renal disease to minimize genetic heterogeneity. SNPs were dropped if HWE p< 1e-5 or any cross-batch association p< 5e-5.

Genetic data was imputed using the Michigan Imputation Server⁸¹ using Minimac3. Imputation was performed using the 1000 Genomes Phase 3 reference panel. Following imputation, SNPs with minor allele frequency greater than 5% were retained, and SNPs with > 5% missing data or evidence of deviation from Hardy Weinberg equilibrium (p< 1 x 10^~4) were removed. SNPs were pruned so that no two SNPS were in linkage disequilibrium (A² > 0.8).

Phenotypic clustering analysis. To stratify SLE patients into clinically relevant clusters, we performed unsupervised clustering on patient phenotypic data. Cluster analysis was performed on the American College of Rheumatology (ACR) criteria and sub criteria. Data were dichotomized to represent absence or presence of each criterion. Multiple correspondence analysis was performed with the PCAmixdata R package⁸². The top two MCA dimensions were retained as selected by the k-fold cross validation scheme implemented in the missMDA R package⁸³⁸⁴. The number of clusters, k, was chosen by maximizing cluster stability measured by Jaccard similarity using a bootstrap resampling based method. Maximum cluster stability was achieved with k= 3 and each cluster had a Jaccard mean stability score greater than 0.82⁸⁵.

Medication use adjustment. Since medication use can modify CpG methylation at certain sites, we aimed to include medications prescribed at the time of blood sampling as covariates in statistical analyses. We performed principal component analysis (PCA) on a dichotomized matrix of current medications at the time of blood sampling for each patient. The top three PCs were chosen using a three-fold cross validation scheme implemented in the missMDA R package⁸³⁸⁴ and included as covariates in subsequent statistical models.

Differential methylation analysis. In order to account for possible confounding due to cell type heterogeneity, we applied the ReFACTor algorithm⁸⁶ implemented in Glint⁸⁷ to infer peripheral blood cell composition. To identify CpG sites associated with clinical clusters, a linear model adjusted for sex, age, cell count estimates, alcohol use, smoking status, genetic ancestry components, and the top three medication principal components was fit using the nestedF mode in the Limma R package⁴⁰. P values were adjusted using the Benjamini Hochberg procedure. All analyses were performed using R version 3.4.2⁸⁸.

Chromatin state enrichment. 15-state chromatin model epigenome data for all human peripheral blood cell types was accessed via the NIH Roadmap Epigenomics Consortium⁴¹. All CpGs on the probe-set were assigned a chromatin state. For each of the 15 chromatin states, a fold statistic was computed using a Fisher’s exact test for enrichment of the chromatin state within the set of cluster-associated CpGs relative to all the CpGs in the probe-set. This process was repeated for H3K4me3, FI3K4me1 , and H3K27ac ChIP seq peaks from the NIH Roadmap Epigenomics Consortium.

CpG race enrichment adjusted for genetic ancestry. Enrichment of race- associated CpGs in the list of differentially methylated CpGs was determined via a permutation method. Briefly, the total number of cluster-associated CpGs (A/ciuster) was obtained for a specified FDR as above. Then, a null set was created by randomly permuting the race labels 1000 times. For each permutation, from the set of cluster-associated CpGs, we computed the number of CpGs associated with the permutated race labels by fitting a linear model for each CpG adjusting for sex, age, cell count estimates, alcohol use, smoking status, the top three genetic principal components, and the top three medication principal components. We then found the number of race associated CpGs in the set of cluster- associated CpGs ( ace) using the same linear model as the null set. We defined an enrichment statistic as the proportion A/_raCe/A/ciuster relative to the mean of the null distribution.

Statistical meQTL analyses. Since the lllumina BeadChip EPIC platform is known to cross-react with several probes if the region contains a SNP⁸⁹, we first removed all probes with SNPs. meQTL analysis was then performed by fitting a linear model adjusted for sex, age, cell count estimates, alcohol use, smoking status, the top three genetic principal components, and the top three medication principal components using the Matrix eQTL R package⁹⁰.

Causal inference test. For each meQTL association, the genotype (G)-methylation (M)-cluster (Y) relationships were assessed using the Causal Inference Test (CIT). To establish a mediation relationship in which genotype acts on the clusters through methylation, four conditions must be satisfied: (1) G and Y are associated, (2) G is associated with M after adjusting for Y, (3) M is associated with Y after adjusting for G, and (4) G is independent of Y after adjusting for M. The CIT p-value is defined as the maximum of the four component test p-values.

External validation. A random forest model was trained on the cluster labels from the study data and cross validation was used to optimize parameters. Methylation data and ACR criteria were obtained from a previously published SLE cohort¹². All ACR criteria were dichotomized in a manner identical to the study data. The random forest model was then applied to the external validation data to generate cluster labels. CpGs that were differentially methylated in the study data were validated using a linear model adjusted for cell composition, genetic principal components, sex, and age. Smoking history was negative for all patients in the validation cohort. Alcohol use and medication principal components were not included in the linear model since these data were not available.

Data availability

DNA methylation, genotype and phenotypic data that support the findings of this study have been deposited in DbGap with the primary accession code of phs001850.v1 p1 . Data is available through an application to a data access committee.

Code availability

All custom code is available at the web address reached by typing github followed by .com/ishanparanjpe/lupus_clustering in search window of a web browser. References

1. Kaul, A. et al. Systemic lupus erythematosus. Nat. Rev. Dis. Prim. 2, 16039

(2016).

2. Urowitz, M. B. et al. Evolution of disease burden over five years in a multicenter inception systemic lupus erythematosus cohort. Arthritis Care Res. (Hoboken). 64, 132-7 (2012).

3. Alarcon, G. S. et al. Systemic lupus erythematosus in three ethnic groups: IX. Differences in damage accrual. Arthritis Rheum. 44, 2797-2806 (2001).

4. Zonana-Nacach, A. et al. Measurement of damage in 210 Mexican patients with systemic lupus erythematosus: relationship with disease duration. Lupus 7, 119-23 (1998).

5. Tan, E. M. et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 25, 1271-1277 (1982).

6. Banchereau, R. et al. Personalized Immunomonitoring Uncovers Molecular Networks that Stratify Lupus Patients. Cell 165, 551-65 (2016).

7. Bradley, S. J., Suarez-Fueyo, A., Moss, D. R., Kyttaris, V. C. & Tsokos, G. C. T Cell Transcriptomes Describe Patient Subtypes in Systemic Lupus Erythematosus. PLoS One 10, e0141171 (2015).

8. Rai, R., Chauhan, S. K., Singh, V. V., Rai, M. & Rai, G. RNA-seq Analysis Reveals Unique Transcriptome Signatures in Systemic Lupus Erythematosus Patients with Distinct Autoantibody Specificities. PLoS One 11 , e0166312 (2016).

9. Contreras, G. et al. Outcomes in African Americans and Hispanics with lupus nephritis. Kidney Int. 69, 1846-51 (2006).

10. Dall’Era, M. et al. The Incidence and Prevalence of Systemic Lupus Erythematosus in San Francisco County, California: The California Lupus Surveillance Project. Arthritis Rheumatol. ( Hoboken , N.J.) 69, 1996-2005 (2017).

11. Gonzalez, L. a, Toloza, S. M. a, McGwin, G. & Alarcon, G. S. Ethnicity in systemic lupus erythematosus (SLE): its influence on susceptibility and outcomes. Lupus 22, 1214-24 (2013).

12. Chung, S. A. et al. Lupus Nephritis Susceptibility Loci in Women with Systemic Lupus Erythematosus. J. Am. Soc. Nephrol. ASN.2013050446 (2014).

13. Mohan, C. & Putterman, C. Genetics and pathogenesis of systemic lupus erythematosus and lupus nephritis. Nat. Rev. Nephrol. 11, 329-341 (2015).

14. Taylor, K. E. et al. Specificity of the STAT4 genetic association for severe disease manifestations of systemic lupus erythematosus. PLoS Genet.4, e1000084 (2008). 15. Lanata, C. M. et al. Genetic contributions to lupus nephritis in a multi-ethnic cohort of systemic lupus erythematous patients. PLoS One 13, (2018).

16. Chung, S. A. et al. Genome-Wide Assessment of Differential DNA Methylation Associated with Autoantibody Production in Systemic Lupus Erythematosus. PLoS One 10, e0129813 (2015).

17. Coit, P. et al. Genome-wide DNA methylation study suggests epigenetic accessibility andtranscriptional poising of interferon-regulated genes in naive CD4+ T cellsfrom lupus patients. J. Autoimmun. 43, 78-84 (2013).

18. Lanata, C. M., Chung, S. A. & Criswell, L. A. DNA methylation 101 : what is important to know about DNA methylation and its role in SLE risk and disease heterogeneity. Lupus Sci. Med. 5, e000285 (2018).

19. Mok, A. et al. Genome-wide profiling identifies associations between lupus nephritis and differential methylation of genes regulating tissue hypoxia and type 1 interferon responses. Lupus Sci. Med. 3, e000183 (2016).

20. Coit, P. et al. Renal involvement in lupus is characterized by unique DNA methylation changes in naive CD4+ T cells. J. Autoimmun. 61, 29-35 (2015).

21. Taylor, K. E. et al. Risk alleles for systemic lupus erythematosus in a large case-control collection and associations with clinical subphenotypes. PLoS Genet. 7, e1001311 (2011).

22. Chung, S. A. et al. Differential genetic associations for systemic lupus erythematosus based on anti-dsDNA autoantibody production. PLoS Genet. 7, e1001323 (2011).

23. De Jager, P. L. etal. Data descriptor: A multi-omic atlas of the human frontal cortex for aging and Alzheimer’s disease research. Sci. Data 5, 180142 (2018).

24. Wu, Y. et al. Integrative analysis of omics summary data reveals putative mechanisms underlying complex traits. Nat. Commun. 9, 918 (2018).

25. Yao, X. etal. Tissue-specific network-based genome wide study of amygdala imaging phenotypes to identify functional interaction modules. Bioinformatics 33, 3250- 3257 (2017).

26. Lyons, J. etal. Integrated in vivo multiomics analysis identifies p21 -activated kinase signaling as a driver of colitis. Sci. Signal. 11, (2018).

27. Shen, R. et al. Integrative subtype discovery in glioblastoma using iCIuster. PLoS One 7, e35236 (2012).

28. Mo, Q. et al. Pattern discovery and cancer gene identification in integrated cancer genomic data. Proc. Natl. Acad. Sci. U. S. A. 110, 4245-50 (2013). 29. Zhang, X. et al. The oral and gut microbiomes are perturbed in rheumatoid arthritis and partly normalized after treatment. Nat. Med. 21, 895-905 (2015).

30. Tasaki, S. et al. Multi-omics monitoring of drug response in rheumatoid arthritis in pursuit of molecular remission. Nat. Commun. 9, 2755 (2018).

31. Ines, L. et al. Classification of Systemic Lupus Erythematosus: Systemic Lupus International Collaborating Clinics Versus American College of Rheumatology Criteria. A Comparative Study of 2,055 Patients From a Real-Life, International Systemic Lupus Erythematosus Cohort. Arthritis Care Res. (Hoboken). 67, 1180-5 (2015).

32. Whitaker, J. W. et al. Integrative omics analysis of rheumatoid arthritis identifies non-obvious therapeutic targets. PLoS One 10, e0124254 (2015).

33. Bello, G. A. etal. Development and validation of a simple lupus severity index using ACR criteria for classification of SLE. Lupus Sci. Med. 3, e000136 (2016).

34. Liberzon, A. et al. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 1, 417-425 (2015).

35. Baechler, E. C. et al. Interferon-inducible gene expression signature in peripheral blood cells of patients with severe lupus. Proc. Natl. Acad. Sci. U. S. A. 100, 2610-5 (2003).

36. Crow, M. K., Kirou, K. A. & Wohlgemuth, J. Microarray analysis of interferon- regulated genes in SLE. Autoimmunity 36, 481-490 (2003).

37. Imgenberg-Kreuz, J. et al. DNA methylation mapping identifies gene regulatory effects in patients with systemic lupus erythematosus. Ann. Rheum. Dis. 77, 736-743 (2018).

38. Coit, P. et al. Epigenome profiling reveals significant DNA demethylation of interferon signature genes in lupus neutrophils. J. Autoimmun. 58, 59-66 (2015).

39. Yang, G. etal. Phosphodiesterase 7A-deficient mice have functional T cells. J. Immunol. 171 , 6414-20 (2003).

40. Ritchie, M. E. et al. limma powers differential expression analyses for RNA- sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).

41. Roadmap Epigenomics Consortium et al. Integrative analysis of 111 reference human epigenomes. Nature 518, 317-329 (2015).

42. Wang, Y. et al. Deubiquitinating enzymes regulate PARK2-mediated mitophagy. Autophagy 11 , 595-606 (2015).

43. Gu, H. et al. Essential role for Gab2 in the allergic response. Nature 412, 186-90 (2001).

44. Adams, S. J., Aydin, I. T. & Celebi, J. T. GAB2-a scaffolding protein in cancer. Mol. Cancer Res. 10, 1265-70 (2012). 45. Ding, C., Yu, W., Feng, J. & Luo, J. Structure and function of Gab2 and its role in cancer (Review). Mol. Med. Rep. 12, 4007-4014 (2015).

46. Gobin, S. J. & van den Elsen, P. J. Transcriptional regulation of the MHC class lb genes HLA-E, HLA-F, and FILA-G. Hum. Immunol. 61 , 1102-7 (2000).

47. Millstein, J., Zhang, B., Zhu, J. & Schadt, E. E. Disentangling molecular relationships with a causal inference test. BMC Genet. 10, 23 (2009).

48. Galanter, J. M. et al. Differential methylation between ethnic sub-groups reflects the effect of genetic ancestry and environmental exposures. Elite 6, (2017).

49. Zhang, Y.-H. etal. Identifying and analyzing different cancer subtypes using RNA-seq data of blood platelets. Oncotarget 8, 87494-87511 (2017).

50. Ramaker, R. C. et al. RNA sequencing-based cell proliferation analysis across 19 cancers identifies a subset of proliferation-informative cancers with a common survival signature. Oncotarget 8, 38668-38681 (2017).

51. Li, P. H. et al. Relationship between autoantibody clustering and clinical subsets in SLE: cluster and association analyses in Hong Kong Chinese. Rheumatology (Oxford). 52, 337-45 (2013).

52. Linnik, M. D. et al. Relationship between anti-double-stranded DNA antibodies and exacerbation of renal disease in patients with systemic lupus erythematosus. Arthritis Rheum. 52, 1129-37 (2005).

53. Richman, I. B. et al. European population substructure correlates with systemic lupus erythematosus endophenotypes in North Americans of European descent. Genes Immun. 11 , 515-521 (2010).

54. Sanchez, E. et al. Impact of genetic ancestry and sociodemographic status on the clinical expression of systemic lupus erythematosus in American Indian-European populations. Arthritis Rheum. 64, 3687-3694 (2012).

55. Richman, I. B. et al. European genetic ancestry is associated with a decreased risk of lupus nephritis. Arthritis Rheum. 64, 3374-3382 (2012).

56. Bennett, L. etal. Interferon and Granulopoiesis Signatures in Systemic Lupus Erythematosus Blood. J. Exp. Med. 197, 711-723 (2003).

57. Renauer, P. et al. DNA methylation patterns in naive CD4+ T cells identify epigenetic susceptibility loci for malar rash and discoid rash in systemic lupus erythematosus. Lupus Sci. Med. 2, (2015).

58. Zhao, M. etal. IFI44L promoter methylation as a blood biomarker for systemic lupus erythematosus. Ann. Rheum. Dis. 75, 1998-2006 (2016).

59. Ivashkiv, L. B. & Donlin, L. T. Regulation of type I interferon responses. Nat. Rev. Immunol. 14, 36-49 (2014). 60. Chen, L. et al. Genetic Drivers of Epigenetic and Transcriptional Variation in Human Immune Cells. Cell 167, 1398-1414.e24 (2016).

61. Garg, P., Joshi, R. S., Watson, C. & Sharp, A. J. A survey of inter-individual variation in DNA methylation identifies environmentally responsive co-regulated networks of epigenetic variation in the human genome. PLOS Genet. 14, e1007707 (2018).

62. Ahsan, M. et al. The relative contribution of DNA methylation and genetic variants on protein biomarkers for human diseases. PLoS Genet. 13, e1007005 (2017).

63. Jucaud, V. et al. Serum antibodies to human leucocyte antigen (HLA)-E, HLA-F and HLA-G in patients with systemic lupus erythematosus (SLE) during disease flares: Clinical relevance of HLA-F autoantibodies. Clin. Exp. Immunol. 183, 326-40 (2016).

64. Lee, N., Ishitani, A. & Geraghty, D. E. HLA-F is a surface marker on activated lymphocytes. Eur. J. Immunol. 40, 2308-2318 (2010).

65. Dulberger, C. L. et al. Human Leukocyte Antigen F Presents Peptides and Regulates Immunity through Interactions with NK Cell Receptors. Immunity 46, 1018— 1029.e7 (2017).

66. Saningong, A. D. & Bayer, P. Human DNA-binding peptidyl-prolyl cis/trans isomerase Pari 4 is cell cycle dependently expressed and associates with chromatin in vivo. BMC Biochem. 16, 4 (2015).

67. Cho, S. H. et al. Glycolytic rate and lymphomagenesis depend on PARP14, an ADP ribosyltransferase of the B aggressive lymphoma (BAL) family. Proc. Natl. Acad. Sci. U. S. A. 108, 15972-7 (2011).

68. Mueller, J. W. & Bayer, P. Small family with key contacts: par14 and par17 parvulin proteins, relatives of pin1 , now emerge in biomedical research. Perspect. Medicin. Chem. 2, 11-20 (2008).

69. lansante, V. et al. PARP14 promotes the Warburg effect in hepatocellular carcinoma by inhibiting JNK1 -dependent PKM2 phosphorylation and activation. Nat. Commun. 6, 7882 (2015).

70. Kono, M. et al. Pyruvate dehydrogenase phosphatase catalytic subunit 2 limits Th17 differentiation. Proc. Natl. Acad. Sci. U. S. A. 115, 9288-9293 (2018).

71. Rockel, J. S. & Kapoor, M. Autophagy: controlling cell fate in rheumatic diseases. Nat. Rev. Rheumatol. 12, 517-31 (2016).

72. Clarke, A. J. et al. Autophagy is activated in systemic lupus erythematosus and required for plasmablast development. Ann. Rheum. Dis. 74, 912-20 (2015).

73. Martinez, J. et al. Noncanonical autophagy inhibits the autoinflammatory, lupus-like response to dying cells. Nature 533, 115-9 (2016). 74. Liu, Q. et al. Broad and diverse mechanisms used by deubiquitinase family members in regulating the type I interferon signaling pathway during antiviral responses. Sci. Adv. 4, eaar2824 (2018).

75. Crow, M. K. Type I interferon in the pathogenesis of lupus. J. Immunol. 192, 5459-68 (2014).

76. Banchereau, J. & Pascual, V. Type I interferon in systemic lupus erythematosus and other autoimmune diseases. Immunity 25, 383-92 (2006).

77. Bronson, P. G., Chaivorapol, C., Ortmann, W., Behrens, T. W. & Graham, R. R. The genetics of type I interferon in systemic lupus erythematosus. Curr. Opin. Immunol. 24, 530-7 (2012).

78. Hochberg, M. C. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 40, 1725 (1997).

79. Maksimovic, J., Gordon, L. & Oshlack, A. SWAN: Subset-quantile Within Array Normalization for lllumina Infinium HumanMethylation450 BeadChips. Genome Biol. 13, R44 (2012).

80. Aryee, M. J. etal. Minfi: A flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays. Bioinformatics 30, 1363-1369 (2014).

81. Hoffmann, T. J. et al. Design and coverage of high throughput genotyping arrays optimized for individuals of East Asian, African American, and Latino race/ethnicity using imputation and a novel hybrid SNP selection algorithm. Genomics 98, 422-430 (2011).

82. Das, S. etal. Next-generation genotype imputation service and methods. Nat. Genet. 48, 1284-1287 (2016).

83. Chavent, M., Kuentz-Simonet, V., Labenne, A. & Saracco, J. Multivariate analysis of mixed data : The PCAmixdata R package. arXiv e-prints 1-31 (2014).

84. Josse, J. & Husson, F. missMDA : A Package for Handling Missing Values in Multivariate Data Analysis. J. Stat. Softw. 70, 1-31 (2016).

85. Josse, J., Chavent, M., Liquet, B. & Husson, F. Handling Missing Values with Regularized Iterative Multiple Correspondence Analysis. J. Classif. 29, 91-116 (2012).

86. Hennig, C. Cluster-wise assessment of cluster stability. Comput. Stat. Data Anal. 52, 258-271 (2007).

87. Rahmani, E. et al. Sparse PCA corrects for cell type heterogeneity in epigenome-wide association studies. Nat. Methods 13, 443-445 (2016). 88. Rahmani, E. et al. GLINT: A user-friendly toolset for the analysis of high- throughput DNA-methylation array data. Bioinformatics 33, 1870-1872 (2017).

89. R Core Team. R Core Team (2017). R: A language and environment for statistical computing. R Found. Stat. Comput. Vienna, Austria. URL http//www.R- project.org/. R Foundation for Statistical Computing (2017).

90. Daca-Roszak, P. et al. Impact of SNPs on methylation readouts by lllumina Infinium HumanMethylation450 BeadChip Array: Implications for comparative population studies. BMC Genomics 16, (2015).

91. Shabalin, A. A. Matrix eQTL: Ultra fast eQTL analysis via large matrix operations. Bioinformatics 28, 1353-1358 (2012).

92. Chen, J., Bardes, E. E., Aronow, B. J. & Jegga, A. G. ToppGene Suite for gene list enrichment analysis and candidate gene prioritization. Nucleic Acids Res. 37, W305-11 (2009).

Accordingly, the preceding merely illustrates the principles of the present disclosure. It will be appreciated that those skilled in the art will be able to devise various arrangements which, although not explicitly described or shown herein, embody the principles of the invention and are included within its spirit and scope. Furthermore, all examples and conditional language recited herein are principally intended to aid the reader in understanding the principles of the invention and the concepts contributed by the inventors to furthering the art, and are to be construed as being without limitation to such specifically recited examples and conditions. Moreover, all statements herein reciting principles, aspects, and embodiments of the invention as well as specific examples thereof, are intended to encompass both structural and functional equivalents thereof. Additionally, it is intended that such equivalents include both currently known equivalents and equivalents developed in the future, i.e., any elements developed that perform the same function, regardless of structure. The scope of the present invention, therefore, is not intended to be limited to the exemplary embodiments shown and described herein.

Claims

WHAT IS CLAIMED IS:

1 . A method of producing a DNA methylation profile of a subject, comprising: assessing DNA methylation in a sample obtained from a subject having or suspected of having systemic lupus erythematosus (SLE), wherein DNA methylation is assessed for a gene or panel of genes comprising one or more of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M, to produce a DNA methylation profile of the subject.

2. The method according to claim 1 , wherein DNA methylation is assessed for a panel of genes comprising two or more, three or more, four or more, five or more, six or more, seven or more, eight or more, or each of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

3. The method according to claim 1 or claim 2, further comprising assigning the subject to an SLE severity subgroup based on the DNA methylation profile.

4. The method according to claim 3, further comprising administering to the subject an SLE therapy indicated for the SLE severity subgroup.

5. The method according to any one of claims 1 to 4, wherein the panel of genes comprises 500 or fewer genes.

6. The method according to any one of claims 1 to 4, wherein the panel of genes comprises 250 or fewer, 150 or fewer, 100 or fewer, 75 or fewer, or 50 or fewer genes.

7. The method according to any one of claims 1 to 6, wherein the sample obtained from the subject is a peripheral blood mononuclear cell (PBMC) sample.

8. The method according to any one of claims 1 to 7, further comprising obtaining the sample from the subject.

9. The method according to any one of claims 1 to 8, wherein the assessing is by bisulphite sequencing.

10. The method according to any one of claims 1 to 8, wherein the assessing is by genotyping CpG sites using a microarray.

11. The method according to any one of claims 1 to 8, wherein the assessing is by genotyping CpG sites using a DNA sequencing system.

12. The method according to claim 11 , wherein the DNA sequencing system is a nanopore-based DNA sequencing system.

13. A method of treating a subject having systemic lupus erythematosus (SLE), comprising: administering to a subject identified as having a DNA methylation profile characteristic of an SLE severity subgroup a therapy indicated for the SLE severity subgroup, wherein the DNA methylation profile comprises the DNA methylation status for a gene or panel of genes comprising one or more of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

14. The method according to claim 13, wherein the DNA methylation profile comprises the DNA methylation status for a panel of genes comprising two or more, three or more, four or more, five or more, six or more, seven or more, eight or more, or each of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

15. The method according to claim 13 or claim 14, wherein the panel of genes comprises 500 or fewer genes.

16. The method according to claim 13 or claim 14, wherein the panel of genes comprises 250 or fewer, 150 or fewer, 100 or fewer, 75 or fewer, or 50 or fewer genes.

17. A method of treating a subject having a disorder characterized by type I interferon signaling, comprising: administering a therapeutic amount of an anti-type I interferon signaling therapy to a subject identified as having a DNA methylation profile characteristic of a disorder characterized by type I interferon signaling, wherein the DNA methylation profile comprises the DNA methylation status for a gene or panel of genes comprising one or more of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

18. The method according to claim 17, wherein the DNA methylation profile comprises the DNA methylation status for a panel of genes comprising two or more, three or more, four or more, five or more, six or more, seven or more, or each of PDE7A, PARP14, IFI44L, MX1 , PARP9, EPSTI1 , RSAD2, IFI27 and B2M.

19. The method according to claim 17 or claim 18, wherein the panel of genes comprises 500 or fewer genes.

20. The method according to claim 17 or claim 18, wherein the panel of genes comprises 250 or fewer, 150 or fewer, 100 or fewer, 75 or fewer, or 50 or fewer genes.

21. The method according to any one of claims 17 to 20, wherein administering an anti-type I interferon signaling therapy to the subject comprises administering an anti- interferon-a agent to the subject.

22. The method according to claim 21 , wherein the anti-interferon-a agent is an antibody.

23. The method according to claim 22, wherein the antibody is sifalimumab.

24. The method according to any one of claims 17 to 20, wherein administering an anti-type I interferon signaling therapy to the subject comprises administering an anti-type 1 interferon receptor agent to the subject.

25. The method according to claim 24, wherein the anti-type 1 interferon receptor agent is an antibody.

26. The method according to claim 25, wherein the antibody is anifrolumab.