WO2005086870A2

WO2005086870A2 - Novel lactobacillus strains and method of use

Info

Publication number: WO2005086870A2
Application number: PCT/US2005/007794
Authority: WO
Inventors: Lin Tao
Original assignee: The Board Of Trustees Of The University Of Illinois
Priority date: 2004-03-10
Filing date: 2005-03-10
Publication date: 2005-09-22
Also published as: WO2005086870A3

Abstract

Disclosed are compositions comprising mannose binding lectin (MBL)-positive Lactobacillus, methods of identifying MBL-positive Lactobacillus, and methods of use.

Description

NOVEL LACTOBACILLUS STRAINS AND METHOD OF USE

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional application no. 60/551,890, filed March 10, 2004, which is incorporated by reference in its entirety.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH

[0002] This invention was made with government support through grant AI50491 and DEI 5499, awarded by the National Institutes of Health. The United States Government has certain rights in this invention.

INTRODUCTION

[0003] Worldwide, more than 40 million people are infected with human immunodeficiency virus (HIV), the causal agent of acquired immune deficiency syndrome (AIDS). As of 2004, approximately 28 million deaths had been attributed to AIDS. The virus is transmitted from host to host through the exchange of bodily fluids, which may occur through sexual contact, blood transfusions, contact with contaminated needles, or breastfeeding.

[0004] Although heterosexual contact is the primary mode of HIV transmission, nearly one million infants become infected with HIV each year. Infection may occur in utero, at birth, or through breast milk. Antiretroviral prophylaxis and delivery by Caesarean section can dramatically reduce the rate of maternal transmission of HIV. However, breastfeeding remains an important route of vertical HIV transmission. It is estimated that from one-third to one-half of infected children acquired HIV by ingesting breast milk containing HIV. Current methods, such as physical barriers and microbicides, are not suitable for use in preventing transmission of HIV to infants through breast milk. [0005] Milk containing HIV is more infective than in other body fluids probably due to larger fluid volume, frequent contact over a relatively large surface area, and/or higher total viral load. Today, for HIV-positive women in more affluent countries, formula feeding is recommended. However, in developing countries that lack access to clean water, breastfeeding is recommended to avoid other health problems (e.g., fatal diarrhea) that can result from feeding infants formula prepared from contaminated water.

[0006] There is an ongoing need for compositions and methods of preventing infection by HIV and other pathogens.

BRIEF SUMMARY OF THE INVENTION

[0007] In one aspect, the present invention provides a method for altering the bacterial composition of the digestive tract or urogenital tract of a mammalian subject comprising administering to the subject bacteria of at least one MBL-positive Lactobacillus strain.

[0008] In another aspect, the present invention provides a composition comprising MBL-positive Lactobacillus bacteria, the bacteria characterized by the ability to co- aggregate with Saccharomyces cerevisiae and to bind to gpl20.

[0009] In yet another aspect, the present invention provides a method of evaluating a Lactobacillus isolate for its suitability for use in methods or compositions of the invention comprising testing the isolate for the ability to co-aggregate with a yeast, bacterium, or virus comprising mannose on its cell wall or envelope, the ability to bind a mannose containing viral coat protein, the ability to bind at least one HIV -target cell, or the ability to block HIV infection. BRIEF DESCRIPTION OF THE DRAWINGS

[00010] Fig. 1 is an image of co-aggregating mannose binding lectin positive Lactobacillus and yeast.

[00011] Fig. 2 is a graph showing the results of quantitative binding of various Lactobacillus isolates to immobilized gpl20.

[00012] Fig. 3 is a graph showing the results of quantitative binding of Lactobacillus fermentum OLB 19a to viral envelop proteins form different clades of HIV and SIV.

[00013] Fig. 4 is an image showing binding of mannose binding lectin positive Lactobacillus to HIV -target cells.

[00014] Fig. 5 is a graph showing the number of HIV viruses bound as a function of the Lactobacillus cell concentration.

[00015] Fig. 6 compares the number of HIV viruses trapped by an MBL-negative Lactobacillus and two MBL-positive Lactobacillus strains.

[00016] Fig. 7 shows the relative fluorescence of Luciferase-tagged HIV present in a cell culture observed following infection with or without preincubation with a MBL- positive Lactobacillus in various concentrations.

[00017] Fig. 8 shows the number of HIV virions in cells exposed to HIV in the presence of Lactobacillus OLB43b as a function of exposure time (Fig. 8A) and cell concentration (Fig. 8B).

[00018] Fig. 9 shows the number of HIV or SIV virions bound by Lactobacillus cells as a function of Lactobacillus strain.

[00019] Fig. 10 is an image showing the presence or absence of capture by MBL positive lactobaciUi to plain green fluorescent polystyrene beads (Fig. 10A) or gpl20- linked green fluorescent polystyrene beads (Fig. 10B). DETAILED DESCRIPTION

[00020] We envision protecting against infection by HIV or other pathogens by enhancing the innate immunity. Innate immunity is the first line of defense, and includes anatomic barriers, defensive cells, humoral factors, and commensal bacteria. Adult humans host a large number of commensal bacteria. Infants begin acquiring commensal bacteria during or after birth. It is envisioned that altering the composition of commensal bacteria of an infant's digestive tract may afford protection against infection by HIV or other pathogens. Similarly, altering the composition of commensal bacteria of the vagina may protect against HIV infection or other pathogens.

[00021] HIV contains mannose in the glycoproteins of its envelope. The presence of mannose in the glycoproteins is fairly constant and relatively insensitive to antigenic shifts in HIV caused by mutational changes. Human mannose-binding lectin (MBL) is a plasma protein that binds to mannose-containing microorganisms, including viruses, bacteria, fungi, and protozoa. Binding of mannose-containing pathogens by MBL facilitates clearance of the pathogen by complement activation and opsonization, and subsequent phagocytosis. However, because human MBL is found principally, if not exclusively, in the plasma, it does not provide infants with a first line of defense at the site of infection, specifically, in the digestive tract.

[00022] As described in the Examples, 170 Lactobacillus strains were isolated from saliva of apparently healthy human subjects and tested for the ability to coaggregate with Saccharomyces cerevisae (baker's yeast), a yeast containing mannose. Of the tested strains, nine were found to co-aggregate with S. cerevisae. The addition of mannose at concentrations of 50 mM or 100 mM prevented or reversed Lactobacillus co-aggregation with S. cerevisae.

[00023] Although the factor or factors responsible for co-aggregation of certain strains of Lactobacillus with S. cerevisae yeast has not yet been fully characterized, evidence gathered to date and described herein below suggest that it is a mannose-binding lectin or lectins. Therefore, Lactobacillus strains having the ability to co-aggregate with S. cerevisae are referred to herein as MBL-positive Lactobacillus strains.

[00024] Transposon mutants of a MBL positive Lactobacillus fermentum strain OLB- 19a were created and a putative MBL disruption mutant was selected on the basis of its inability to bind yeast, and designated TM-4A3. The region of the chromosome of TM- 4A3 into which the transposon was inserted has been subcloned and sequencing of that region of the putative MBL disruption mutant is currently under way.

[00025] As described in the Examples, the nine Lactobacillus strains capable of co- aggregating S. cerevisae were evaluated for the ability to bind to the HIV envelope glycoproteins gpl20 from strain SF2. Six strains (OLB- 12, OLB- 19a, OLB-24b, OLB- 36b, OLB-43b, and 101S) were found to bind to gpl20. The mutant TM-4A3 does not bind to gpl20. OLB- 19a, which was found to bind to gpl20 with a relatively high affinity, was further tested and found to bind to envelope glycoproteins from several different clades of HIV or simian immunodeficiency virus (SIV). The six gpl20-binding strains were assigned to three different species based on 16S rRNA gene analysis as follows: OLB-19a and 101S belong to L. fermentum; OLB-24b belongs to L. gallinarium; and OLB- 12, OLB-36b, and OLB-43b belong to L. delbrueckii subsp. lactis.

[00026] In the Examples below, binding of Lactobacillus to HIV envelope glycoproteins was measured by two different means. In a quantitative method, the glycoprotein of interest was immobilized in the wells of a microtiter plate, the glycoprotein was contacted with cells of a test bacterial strain, unbound cells were removed by washing, bound cells were removed by elution with mannose, and the cells were plated and counted. In another example, the HIV glycoprotein was immobilized on a glass slide, contacted with cells of a test bacterial strain, washed and Gram-stained to visually detect bacteria bound to the slide through the glycoprotein. It is envisioned that any suitable means may be used in screening Lactobacillus isolates for the ability to bind to HIV envelope glycoproteins. [00027] As described in the Examples below, virus capture ELISA demonstrated that both Lactobacillus OLB- 19a and OLB-43b bind to free HIV-l_Ba virus in a dose- dependent manner, and that OLB- 19a binds more free HIV than OLB-43b binds.

[00028] In addition to binding free virus, OLB- 19a was found to bind all HIV target cell types tested, including T and B lymphocytes, monocytes, and cells expressing DC- SIGN. The binding was reversible upon addition of 50 mM mannose.

[00029] Despite the apparent ability of OLB-19a cells to bind HIV virus with a higher affinity than OLB-43b, OLB-43b affords greater protection against infection by HIV in a dose-dependent manner. The protective effect may be due to factors that destroy HIV viruses, such as secreted or surface-bound MBL or proteinases.

[00030] As further described in detail in the Examples, a number of vaginal MBL positive Lactobacillus isolates were characterized with respect to various properties, including yeast binding, gpl20 binding, HIV trapping, and inhibiting the growth of yeast or pathogenic bacteria. It is envisioned that isolates having desirable characteristics may be used to alter the composition of bacteria present in the vagina. This may afford protection against viruses such as HIV, yeast, or pathogen bacteria. The protection may be transient or extended over time, depending on the length of time the MBL positive Lactobacillus strains persist.

[00031] Although it is envisioned that a single MBL positive Lactobacillus strain used alone will afford protection against pathogens, it is envisioned that more than one MBL positive Lactobacillus strain may be used in combination. Suitably, such combinations may include a Lactobacillus strain or strains having a surface bound MBL and a Lactobacillus strain that secrets MBL. For example, a strain such as OLB- 19a may be used in combination with a strain such as OLB-43b.

[00032] Oral Lactobacillus strains producing MBL have been shown to reduce the ability of HIV contacted with the strains to infect HIV target cells. The ability of a MBL positive Lactobacillus strain to reduce infectivity of HIV may be associated with its mannose binding activity. It is therefore reasonably expected that the Lactobacillus strains will also afford protection against or reduce colonization of the digestive tract or infection by other mannose or MBL-like adhesin containing organisms, including, but not limited to, pathogenic E. coli, Salmonella species, Vibrio cholerae, Pseudomonas aeruginosa, Klebsiella pneumoniae, Enterobacter cloacae, Candida species, Aspergillus fumigatus, Cryptosoporidium parvum, Neisseria meningitidis, Neisseria gonorrhoeae, Chlamydia trachomatis, rotaviruses, hepatitis B, hepatitis C, herpes simplex, Nipah (a causal agent of fatal hemorrhagic encephalitis), Εbola, and influenza A and B viruses.

[00033] Although all of the Lactobacillus strains described in the Examples are of oral or vaginal origin, it is specifically envisioned that Lactobacillus strains isolated from other sources (e.g., milk) would be useful in the practice of the present invention.

[00034] We have identified potentially useful putative MBL producing oral and vaginal Lactobacillus isolates, of which two have been characterized in greater detail, and five putative MBL producing vaginal Lactobacillus isolates. It is envisioned that there exist other Lactobacillus strains suitable for use in the present invention that may be identified by screening using any suitable assay, including, but not limited to, those described herein. For example, a Lactobacillus isolate can be evaluated for putative MBL activity using coaggregation assays, binding to a mannose-containing glycoprotein, or by binding to yeast. The ability to bind to HIV target cells or to free viruses, or to protect HIV target cells, can be assayed as described herein, or using any other suitable assay.

[00035] Two strains according to the present invention, Lactobacillus fermentum OLB- 19a and L. delbureckii subsp. lactis. OLB-43b, were deposited under the terms of the Budapest treaty with the American Type Culture Collection in Manassas, VA on March 5, 2004 as ATCC accession numbers PTA-5849 and PTA-5850, respectively. One of skill in the art would appreciate that these strains may be acquired and used as is or may be used to select, mutagenize, genetically engineer, or otherwise develop derivatives that may also be useful in the practice of the invention. Accordingly, the present invention includes derivatives of Lactobacillus fermentum OLB- 19a and L. delbureckii subsp. lactis. OLB-43b. As used herein, a derivative of a Lactobacillus strain is an organism that is selected, mutagenized, genetically engineered or otherwise obtained from that strain.

[00036] For use with infants, it is envisioned that an MBL positive Lactobacillus strain or strains may be delivered as a probiotic shortly after birth to an infant at risk of orally contracting HIV or other pathogens containing mannose or MBL-like adhesins through breast milk or contaminated water or any other means. Human T-cell lymphotropic virus can also be transmitted from mother to child through breast milk, including colostrum, presumably through T-lymphocytes infected with the virus present in breast milk. It is likely that hepatitis may also be present in breast milk of infected women. Prior to administration, the bacteria may be lyophilized and stored in the lyophilized form, and reconstituted just prior to administration with an acceptable carrier, such as sterile water or breast milk, and delivered in an amount effective to establish colonization.

[00037] The Lactobacillus may be delivered as a symbiotic in conjunction with one or more prebiotics. Prebiotics are carbohydrates that are used probiotic bacterium, but are not substantially used by the host or other bacteria present. Symbiotics are a preferred formulation, because the added prebiotic sugar or sugars can promote the growth and colonization of the probiotics. Because trehalose can be used by many bacteria, it is not suitable as a prebiotic. Glucono-delton-lactone (GDL) can be utilized by OLB 19a, and mannitol can be utilized by OLB43b, but neither can be used by the host and by most other bacteria. Therefore, these two sugars are prebiotics for these lactobaciUi. They can be added to the freeze-dried Lactobacillus powder at a suitable ratio (e.g., 10:1 w/w) to facilitate their revival and colonization in the host. GDL is a natural constituent of many foods and mannitol is an edible food additive. Both GDL and mannitol are generally regarded as safe (GRAS) by FDA.

[00038] It is well within the ability of one skilled in the art to develop a protocol for establishing colonization of an infant digestive tract by a Lactobacillus strain. For example, colonization could be achieved using a modification of the method described in Agarwal et al. (J. Pediatric Gastroenterologv and Nutrition 36:397-402, 2003), which is incorporated by reference herein. Agarwal et al. describes the delivery of Lactobacillus rhamnosus strain GG to low birth weight neonates for use as a probiotic. Agarwal et al. describes using twice daily doses of 10⁹ organisms for eight days for newborns receiving antibiotics. It is envisioned that fewer doses may be necessary for infants not on antibiotics. Colonization may be evaluated by performing colony counts on stool samples. It is envisioned that it would be preferable to deliver the bacteria to the infant as soon as possible after birth. Infants receiving the bacteria later may require more doses to effectively colonize the digestive tract. An infant may require an additional dose or doses following a course of antibiotics.

[00039] In addition to protecting infants against infection by orally ingested mannose or MBL-like adhesin containing pathogens, it is envisioned that MBL-producing Lactobacillus strains may be used in a subject of any age to colonize the digestive tract, vagina or rectum and may serve as a barrier to infection by mannose or MBL-like adhesin containing pathogens through those routes. Colonization of the vagina may be especially useful in protecting against infection by HIV, hepatitis B, hepatitis C, herpes simplex, E. coli, Neisseria gonorrhoeae, and Chlamydia trachomatis, or in preventing or treating bacterial vaginosis or vaginal candidiasis.

[00040] The Lactobacillus may be provided in any suitable form, including, but not limited to, lyophilized form as a powder for direct ingestion or for ingestion following reconstitution, as a powder spray, as nipple cream, in capsule form, as a rectal or vaginal suppository, in feminine hygiene products such as tampons, as a candy-like product, as a nasal spray, or in food, such as yogurt. The Lactobacillus may be provided as a pharmaceutical composition with a pharmaceutically acceptable carrier or excipient.

[00041] It is envisioned that the present invention would be suitable for use in economically important species other than humans, including livestock, or in species that serve as reservoirs of disease. For example, diarrhea in cattle caused by mannose or MBL-like adhesin containing pathogens may be reduced by colonizing the gut of cattle with an MBL-positive Lactobacillus. The Lactobacillus strains could be delivered in liquid form by reconstituted lyosphilized bacteria, in a capsule, or mixed with feed.

[00042] Delivery of an isolated MBL protein from a Lactobacillus strain to the nasal passages may confer protection against or treat infections acquired through that route or respiratory tract or lungs infections, such as influenza, Klebsiella, or Pseudomonas. The putative cloned MBL coding sequence will allow construction of an expression vector that will facilitate isolation of the MBL polypeptide sequence, which can be purified for nasal or oral delivery to subjects at risk of infection or infected with influenza or diarrhea caused by mannose or MBL-like adhesin containing pathogens. The purified MBL could be delivered vaginally before or after intercourse to inactivate mannose or MBL-like adhesin containing pathogens associated with sexually transmitted disease.

[00043] The following non-limiting examples are intended to be purely illustrative.

EXAMPLES

[00044] Isolation of Oral Flora

[00045] Saliva was aseptically collected from healthy human subjects and plated on the Lactobacillus Rogosa (Difco) agar plates and incubated at 37°C for 24-48 h anaerobically. For liquid cultures, a single colony was transferred to Lactobacillus MRS broth incubated at 37°C for 16-24 h. When required, the culture was stored at -70°C with 15% glycerol.

[00046] Coaggregation of Lactobacillus strains with Saccharomyces cerevisiae

[00047] Lactobacillus cells (10⁷/10μl) from cultures of late log or stationary phase were washed with phosphate-buffered saline (PBS) and combined on a glass slide with Saccharomyces cerevisiae (10⁶/10μl) grown in YPD broth at 37°C for 24 h and harvested similarly by resuspending in PBS. The bacterium and yeast mixture was incubated for 10 min at room temperature. The cells were observed under a light microscope to detect the presence or absence of coaggregation of the bacteria and yeast cells. Representative results obtained with Lactobacillus fermentum OLB- 19a are shown in Fig. 1. Panel A shows OLB- 19a alone, panel B shows S. cerevisiae alone, panel C shows OLB- 19a and S. cerevisiae, and panel D shows a OLB- 19a and S. cerevisiae in the presence of mannose (50 mM). Lactobacillus fermentum OLB- 19a causes clumping of yeast cells (Fig. 1C), which does not occur in the presence of 50 mM mannose (Fig. ID).

[00048] Plasmid-mediated mutagenesis of OLB-19a

[00049] The Gram-positive temperature-sensitive transposon-like plasmid pGh9:ISS7 (Maguin et al, 1996) was used to mutagenize the MBL⁺ strain L. fermentum OLB- 19a. First, the plasmid was transformed into OLB- 19a by electroporation (Wei et al, 1995). The transformants were selected on MRS agar plates supplemented with 5μg/ml erythromycin at 28°C for 48 - 72 h in a candle jar. The transformants were tested for plasmid integration rate by replica-plating with one plate incubated at 42°C and the other at 28°C. A transformant displayed a 1% plasmid integration rate at 42°C was selected for further studies. Mutagenesis was achieved by shifting a mid-exponential phase culture (10 ml) from 28°C to 42°C for 2 h without antibiotic selection, and subsequently plating on MRS agar plates supplemented with 5μg/ml erythromycin and incubating at 37°C.

[00050] Identification of putative MBL knock-out mutant

[00051] About 3,000 plasmid insertion mutants were screened by a dot-blot yeast- binding method. Baker's yeast was first washed in PBS and fixed with 2% glutaraldehyde solution for 1 h at 37°C. The fixed yeast cells were washed twice with PBS and resuspended in 10 mg/ml glycine for 30 min at room temperature. The cells were washed again twice with PBS and resuspended in PBS supplemented with 0.02% sodium azide. The staining was performed by adding 30 μl 5% Safranin (in 95% ethanol) to 1 ml of yeast cells and incubate for 5 min at room temperature. The red dye-stained yeast cells were extensively washed with PBS until the supernatant became clear. Mutant Lactobacillus clones were grown overnight in MRS broth in microtiter plates and transferred to nitrocellulose membranes. The bacterial cells were dried at room temperature and soaked in PBS. The red dye-stained yeast was used as a probe to react with the bacterial blot on the nitrocellulose membranes in a plastic pouch for 1 h on a rocker at room temperature. The membrane was extensively washed with PBS to remove unbound yeast. Among all tested mutants, only one, Tm-4A3, failed to bind the red dye- stained yeast. This mutant was used as a negative control isogenic MBL-negative strain for subsequent experiments. The plasmid-tagged MBL gene was cloned in E. coli by isolating the mutant DNA, self-ligation, transformation of E. coli and selection of erythromycin resistant transformants at 28°C. The DNA sequence of the cloned pGh9:ISS7 inserted gene is currently being determined.

[00052] Binding of Lactobacillus isolates to HIV envelope glycoproteins

[00053] Lactobacillus strains that co-aggregated with yeasts were tested for their binding to HIV envelope proteins. First, the envelope protein (NIH AIDS Research and Reference Reagent Program, Cat# 386) of one representative strain, HIV-1_SF2, was used for initial selection. Overnight cultures of nine oral Lactobacillus strains and ten vaginal strains in MRS were harvested by centrifugation and washed twice with PBS and resuspended in PBS. The cells were subjected to brief sonication by a microsonic cell disrupter (Komtes, Model KT50) to break clumps. A 100 μl cell suspension (about 10 cells) was added, in triplates, to each well of microtiter plate pre-coated with 0.2 μg HIV- 1SF₂ gpl20. The wells pre-coated with porcine gastric mucin were used as a negative control glycoprotein. The isogenic MBL-negative mutant Tm-4A3 was used as a negative control Lactobacillus strain. The microtiter plates were incubated for 20 min at room temperature. The unbound cells were removed by washing with PBS five times. The bound cells were eluted with 50 mM mannose and plated on MRS agar with serial dilutions for colony counts. Binding was also tested on protein coated glass slides.

[00054] The results indicate that of the nine tested MBL positive strains, six (OLB- 12, OLB-19a, OLB-24b, OLB-36b, OLB-43b, and 101 S) bound to gpl20 with greater affinity than to mucin, and OLB-19a, OLB-36, and OLB-43b appeared to bind to gpl20 with the highest affinity. Five of the ten MBL positive vaginal Lactobacillus strains bound to gpl20. In contrast, the MBL-knockout transposon mutant TM-4A3 did not bind to gpl20 with greater affinity than to mucin (Fig. 2).

[00055] Binding of Lactobacillus to HIV envelope glycoproteins from various HIV clades

[00056] The MBL positive Lactobacillus strain OLB- 19a and its isogenic MBL- knockout mutant TM-4A3 were tested for the ability to bind to HIV envelope glycoproteins from various HIV or simian immunodeficiency virus (SIV) clades on microtiter plates, essentially as described in the previous section, or on glass slides that were first coated with protein, reacted with lactobaciUi, washed, and visualized by Gram staining. The viral envelope proteins were all obtained from NIH AIDS Research and Reference Reagent Program, including gpl20s of HIV_CM235 (Cat# 2698), HIV_BaL (Cat# 4961), HIV_SFI₆2, (Cat# 7363) HIV_SF2 (Cat# 386), HIV.,™ (Cat# 4683), HIV_Cι_adeE (Cat# 3234), and HIV_CN54(Cat# 7749); gpl30 of SIV_MaC239 (Cat# 2322); and gpl40 of SIV_Maci_Aii (Cat# 2209).

[00057] The results indicate that OLB- 19a was able to bind to all tested HIV-gpl20, regardless of the clade from which it was obtained, as well as to SIV glycoproteins gpl30 and gpl40 (Fig. 3). In parallel experiment, Lactobacillus isolates were evaluated for the ability to bind to envelop proteins from different HIV or SIV clades deposited onto a glass slide. If was found that binding of MBL-positive Lactobacillus could be detected in a qualitative slide assay.

[00058] Lactobacillus co-aggregation with HIV-target cells

[00059] Because HIV can be transmitted by cell-free or cell-associated viruses, we evaluated the ability of the MBL-positive Lactobacillus strains to bind to the cell-free or cell-associated forms. We postulated that if HIV-infected cells, such as the CD4 T- lymphocytes, in milk or semen, co-aggregate with bacteria, the cells will be trapped, thus reducing contact with HIV-target cells and cell-associated viral infection. [00060] Three oral Lactobacillus strains (OLB- 19a, OLB-43b and Tm-4A3) were tested for co-aggregation with cell lines representing different HIV-target cells, including T lymphocytes (HPB-ALL-CD4, Jurket, CEMl, and CEM2), B lymphocytes (Jy), and monocytes (THP-J, THP-DC-SIGN, and U937), with and without 10 mM CaCl and/or 50 mM mannose. The cells were cultured in RPMI-1640 medium (MediaTech, Cat# 10-040- CV). Cells from overnight cultures of Lactobacillus strains in MRS were pelleted, washed and resuspended in the RPMI-1640 medium such that the cells were concentrated 10-fold. 10-μl aliquots of bacterial and mammalian cells were mixed on a glass slide and incubated at room temperature for 10 min. The cell mixture was then observed under a light microscope for formation of co-aggregations.

[00061] Of the tested strains, only OLB- 19a co-aggregated with all the HIV-target cells tested. OLB-43b and the MBL mutant strain Tm-4A3 did not co-aggregate with any of these cells. The co-aggregation between OLB— 19a and these HIV-target cells was blocked or reversed by 50 mM mannose or alpha-D-mannoside.

[00062] Virus capture p24 ELISA

[00063] Two MBL-positive strains, L. fermentum OLB- 19a and L. delbueckii OLB- 43b, were tested for virus binding, and the isogenic MBL mutant strain L. fermentum Tm- 4A3 was used as a negative control. The Lactobacillus strains were grown overnight at 37°C in 20 ml MRS medium each. The bacterial cells were harvested by centrifugation and washed three times with the cell culture RPMI-1640 medium and resuspended in the same medium at 10% of the original volume. In the case of OLB- 19a, the cells were subjected to brief sonication to break up clumps by a microsonic cell disrupter (Komtes, Model KT50). The cells were maintained on ice for about 1-2 h before mixing with HIV viruses. Different concentrations of HIV-lβ_aL viruses (up to 2.1 x 10 /virion 50 μl) and Lactobacillus cells were mixed in the cell culture medium (without antibiotics) and incubated at 37°C for 1 h. The bacterial cells were centrifuged for 5 min at 3000 xg and the pellet were assayed for bound HIV by the virus-capture p24 ELISA (Perkin Elmer, Cat# NEK050001KT). [00064] The number of bound HIV was plotted as a function of the number of OLB- 19a cells (Fig. 5).

[00065] The number of virions bound is shown as a function of cell type (Fig. 6A), with OLB- 19a binding more than OLB-43b.

[00066] Blocking of HIV infection by Lactobacillus

[00067] The unbound viruses present in the supernatants after centrifugation following virus capture (see previous Example) were used to infect PHA-stimulated peripheral blood mononuclear cells (PBMC) at 37°C in a CO₂ incubator. After 4 days of incubation, the infected cells were assayed by p24 ELISA for the blockage of HIV infection in comparison with viruses without mixing with lactobaciUi.

[00068] After 4 days of incubation, the infected cells were assayed by p24 ELISA for HIV. Although OLB- 19a cells were found to bind more HIV viruses, OLB-43b afforded greater protection of cells from infection with HIV (Fig. 6B). Increased HIV killing by OLB-43b may be due to secreted MBL and/or other potential viricidal acitivities, such as proteinase, acid and hydrogen peroxide.

[00069] Blocking of luciferase-tagged HIV by Lactobacillus

[00070] Luciferase-tagged HIV can be used to monitor the rate of HIV infection because the infected cells will express the enzyme luciferase and will glow. The 293T cells were co-transfected with the pHXB2 and pNL4-3.Luc.R-E- plasmids (catalog # 1069 and 3418; NIH AIDS Research and Reference Reagent Program) to produce luciferase-tagged HIV-l_Hχ_B2 viruses, which were used to infect U87 (U87.CD4.CXCR4: catalog# 4036) receptor cells after interacting with lactobaciUi. Briefly, 293T cells were seeded to 4x10 cells per 100 mm plate the day before transfection. Calcium phosphate precipitation was then used to co-transfect the 293T cells with the pHXB2 plasmid and the pNL4-3.Luc.R-E- plasmid. The transfection was allowed to progress for 48 h, after which, the medium from the 293T cells was harvested and filtered through a 45 μm filter to make the virus particle stock. [00071] The day before infection, U87 cells were seeded to lxlO⁵ cells/well of a 12- well cell culture plate in a volume of 1 ml. The following day, 200 μl of the virus particle stock was added to each of the wells of the U87 cells after removal of the medium. The plates were incubated overnight at 37°C in a CO₂ incubator. After approximately 16 h, the virus was aspirated and replaced with U87 medium (Dulbecco's modification of Eagle's medium, Cat# 10-013-CV, MediaTech) and the cells were allowed to rest for another 24 h. The luciferase activity was measured in triplicate using the Luciferase Assay System from Promega (cat# E4030) and a Berthold FBI 2 luminometer running Sirius software.

[00072] The lactobaciUi were grown 37°C in MRS broth overnight to reach stationary phase. The cells were washed with the receptor cell U87 medium and resuspended in 10% of its original volume (about 10⁹ cells/ml). OLB- 19a cells were subjected to a brief sonication to break up clumps. The viruses (1 :16 dilution) and the MBL+ strains OLB- 19a, OLB-43b, or the MBL mutant strain Tm4A3 were mixed in equal volumes with the bacteria being diluted to varying concentrations (1 :2 through 1 :8) in the receptor cell media. This medium contained 100 unit/ml penicillin G and 200 μg/ml streptomycin, which prevented growth of the lactobaciUi. These samples of virus and bacteria were rocked gently at room temperature for 20 minutes and incubated on the bench for another 40 minutes. The samples were then centrifuged at 10,000 x g for 3 minutes and filtered through a 0.45 μm filter, added at 400 μl to the receptor cells, which was seeded the day before in a 12-well-plate to lxlO⁴ cells per ml, and incubated at 37°C in a CO₂ incubator overnight. The following morning, new media was added to the cells. After 24 h continued incubation at 37°C in a CO₂ incubator, the cells are lysed and assayed for luciferase activity (Luciferase assay kit E4030, Promega, Madison, WI) according to the manufacturer's instructions. The luciferase readout was performed in triplicate at a tenfold dilution of receptor cell lysate on a triplicate set of infected receptor cells. The average of the luciferase triplicate readout was used as the output for each of the experiments and the triplicate of these experiments is used to calculate the standard deviation. [00073] The results of inhibition of HIV infection by Lactobacillus strains are summarized in the bar graph shown in Fig. 7. The Y axis shows the relative light units, which is correlated with HIV infection. A comparison of a control infection (the first bar) is virus alone without the addition of bacteria.. OLB- 19a and mutant TM4A3 did not afford significant protection of the cells against the free virus. In contrast, OLB-43b afforded dose-dependent protection of cells against HIV even at the highest tested dilution (1:8).

[00074] The ability of OLB-43b cells to protect PBMC against HIV_BaL in a dose- dependent manner was further evaluated with the methods described above using from 0- 3.2 x 10⁹ OLB-43b cells and 1.26 x 10⁹ viral particles at initial contact. Fig. 8 show the relationship between HIV infection and the number of OLB-43b cells that contacted the virus. The viral numbers in the figure represent amplification by PBMC cells for 4 days after infection. Under these conditions, killing approached 100% with 3.2 x 10⁹, while 10⁸ cells afforded 92% killing.

[00075] A time course study in which the virus and bacteria were incubated for different lengths of time was performed. The results (Fig. 9) show the relationship of HIV infection as a function of contact time between L. delbrueckii OLB-43b and HIV- lβ_aL- Under these conditions, maximal killing occurred after just 20 minutes incubation.

[00076] Virus capture by p27 ELISA

[00077] Two MBL-positive strains, L. fermentum OLB- 19a and L. delbueckii OLB- 43b, were tested for virus binding, and the isogenic MBL mutant strain L. fermentum Tm- 4A3 was used as a negative control. The Lactobacillus strains were grown overnight at 37°C in 20 ml MRS medium each. The bacterial cells were harvested by centrifugation and washed three times with the cell culture medium RPMI-1640 and resuspended in the same medium at 10% of the original volume. In the case of OLB 19a, the cells were subjected to brief sonication to break up clumps by a microsonic cell disrupter (Komtes, Model KT50). The cells were maintained on ice for about 1-2 h before mixing with HIV or SIV viruses. Different concentrations of HIV-2 or SIV viruses (up to 2.1 x 10⁸/virion/50 μl) and Lactobacillus cells were mixed in the cell culture medium (without antibiotics) and incubated at 37°C for 1 h. The bacterial cells were centrifuged for 5 min at 3000 xg and the pellet were assayed for bound HIV-2 or SIV by the virus-capture p27 antigen ELISA using RETRO-TEK SIV (Zeptometrix, Catalog # 0801169).

[00078] The number of bound HIV-2 or SIV was plotted as a function of cell type (Fig. 10).

[00079] Binding of MBL positive Lactobacillus to mock HIV particles

[00080] To visualize the capture of HIV particles by lactobaciUi, a mock HIV model was constructed by cross-linking HIV-1_SF2 gpl20 to green fluorescent polystyrene beads (0.8 μm) (Spherotech, Inc., Libertyville, Illinois, USA) and mixed with isopropanol- killed, red fluorescence-stained L. fermentum OLB 19a cells (Live/Dead Stain, Molecular Probes, Eugene, Oregon, USA). A fluorescent Olympus camera microscope was used to observe capture of mock HIV by red fluorescence stained lactobaciUi.

[00081 ] Isopropanol-killed L. fermentum OLB 19a cells did not bind plain beads (Fig. 10A), but bound mock HIV particles (Fig.10B). Individual OLB 19a cells appeared to capture multiple mock HIV particles along the entire cell surface, suggesting that the Lactobacillus MBL is localized to the bacterial cell surface, rather than on a fimbriae-like structures, as in E. coli.

[00082] Blocking HIV-1 in vitro infection by vaginal lactobaciUi

[00083] To identify HIV-capturing strains among vaginal lactobaciUi for use in preventing infection through sexual intercourse, 800 vaginal lactobacillus isolates were screened for co-aggregation of yeast as described above, and 26 isolates were found to bind yeast. Those 26 isolates were evaluated essentially as described above for the oral lactobacillus isolates. Twelve were found to adhere to gpl20-coated plates. These strains were subsequently tested for ability to block HIV-1 infections and to coaggregate with HIV-target cells, CD4+ ALL T lymphocytes and DC-SIGN+ monocytes as described above. Eight strains were found to block HIV-1 infection. Three strains coaggregated with both DC-SIGN and CD4 cells, and two coaggregated with only CD4 cells. The ability to co-aggregate with HIV-target cells indicates that these bacteria can block cell-associated HIV. Therefore, the five strains (A, KC55b; B, KC86a; E, Chilόa; G, Chil39a; and I, Chi895a) that both blocked free HIV-1 and co-aggregated with HIV- target cells are suitable strains for the development of a probiotic microbicide.

[00084] Growth inhibition of vaginal bacterial and yeast pathogens

[00085] The five vaginal strains and six oral strains of Lactobacillus were further evaluated for the ability to protect against infection by bacterial vaginosis-associated bacteria (Esherichia coli ATCC14243, Staphylococcus aureus ATCC25923, Pseudomonas aeruginosa ATCC27853, Gardnerrella vaginalis ATCC14018, Gardnerrella vaginalis ATCC49145, and Mobiluncus curtisii subsp curtisii ATCC35241), or yeast associated with vaginal infections (Candida albicans ATCC 10231, C. krusei ATCC 14242, C. glabrata ATCC66032, C. tropicalis ATCC 13803, C. kefir ATCC46764, and C. parapsilosis ATCC22019). Inhibition was tested independent of mannose binding on agar plates. When grown in broth, the MBL+ lactobaciUi captured the yeast by mannan binding, and the killing rate was 100% against all six tested Candida species.

[00086] The ability of Lactobacillus OLB 19a to inhibit fungal growth was further evaluated as follows. The Lactobacillus OLB 19a culture was streaked in two parallel lines on a square MRS plate and incubated overnight at 37°C. A MBL negative strain, Lactobacillus OLB21a, was substituted for Lactobacillus OLB 19a in a parallel experiment. Cultures of each of the six Candida isolates were spotted between the two Lactobacillus streaks, and at the far side of the plate, and incubated overnight. Inhibition of the yeast was observed by comparing the yeast drops in the two places on the same agar plate. The percentage inhibition was estimated by visually comparing yeast growth at the two sites at which the cultures were spotted. The results indicate that Lactobacillus OLB 19a dramatically inhibits growth of five of the six tested Candida species, whereas Lactobacillus OLB21a did not noticeably inhibit growth. The results suggest that the factor or factors responsible for growth inhibition is secreted and diffusable.

[00087] Lyophilization of Lactobacillus and evaluation of viability

[00088] Lyophilized preparations of select Lactobacillus strains were made as follows. Each bacterial strain was grown in 500 ml MRS broth to the stationary phase (about 16 h). The cells were harvested and washed once with phosphate-buffered saline, repelleted, and the pellet was resuspended in 5 ml 10% trehalose, sucrose, cellobiose, lactose or glucono-delton-lactone (GDL). The cell suspension was frozen at -80°C for 1 hour in a lyophilizing bottle. The freeze-drying was conducted with a Labconco Lyophilizer for about 20 h.

[00089] Following lyophilization, the lyophilized samples were stored at 4°C, 20°C, or 37°C for different periods of time. Revival rates of the freeze-dried bacteria were evaluated by colony counts following rehydration. The results indicate that all five sugars afforded some protection. OLB 19a is protected best by trehalose, while OLB43b by GDL. Superior protection by GDL was particularly evident at higher temperatures (37°C). Neither strain fermented its cryoprotective sugar. More than 30% of all cells survived the freeze-dry process. It is possible that these non-fermentable sugars could offer better protection. With extended storage for 140 days, the bacterial survived best at 4°C (about 10%), followed by 20°C (about 1%), and 37°C (about 0.1%). Because nonviable lactobaciUi effectively captured mock HIV upon contact (Fig.10B), Lactobacillus MBL protein functions independently of its bacterial cell viability. Therefore, a lyophilized culture could offer immediate protection even if most of its bacterial cells are nonviable. However, for long-term protection, it is important that at least some of the lyophilized cells are viable because only live bacteria can grow and colonize the host. T

Species Strain Prebiotic Vancomycin H₂0₂ Anti-bacterial activity Antifunqal activity (% Sugar MIC (μg/ml) Ec Sa Pa Gv1 GV2 Mc 1 2 3 4 5 6

O l isolates:

L dslbrueckii subsp lactis OLB12 Mannitol 1 + + + + 3+ - 3+ 100 80 100 100 100 100

L dslbrueckii subsp lactis OLB36b Mannitol 1 + + + + 2+ - 3+ 100 80 100 100 100 100

L dslbrueckii subsp lactis OLB43b Mannitol 1 + + + + 3+ 3+ 3+ 100 80 99.9 100 100 100

L gallinarum OLB24b GDL >1 ,250 - - - - + - - 100 80 99.9 100 99.9 100

L fermentum 0LB19a GDL 1 ,000 - - + - 2+ 3+ 3+ 100 80 99.9 100 99.9 100

L fermentum 101S GDL >1 ,250 - - - - + - - 100 80 99.9 100 99.9 100

Va inal isolates:

L fermentum KC55b GDL 1 ,000 - - + + - - - 100 80 99.9 100 99.9 100

L fermentum (cellobiosus) KC86a GDL >1 ,250 - - + + - - - 100 80 99.9 100 100 100

L fermentum Chi139a GDL >1 ,250 - - + - + - - 100 80 99.9 99.9 99.9 100

L reuteri Chi16a GDL 500 - - - - 3+ 2+ - 100 80 99.9 99.9 99.9 100 I

L vaginalis Chi895a GDL 500t - - + - - - - 100 80 99.9 99.9 99.9 100 r I

¹ Ec, Esherichia coli ATCC14243; Sa, Staphylococcus aureus ATCC25923; Pa, Pseudomonas aeruginosa ATCC27853; Gvl, Gardnerrella vaginalis ATCC14018; Gv2, Gardnerrella vaginalis ATCC49145; Mc, Mobiluncus curtisii subsp curtisii ATCC35241. 3+, Strong, 2+, moderate; +, weak; +, borderline; -, negative. The antimicrobial activity was analyzed with the soft-agar overlay method.

² Candida albicans ATCC10231; 2, C. krusei ATCC14242; 3, C. glabrata ATCC66032; 4, C. tropicalis ATCC 13803; 5, C. kefir ATCC46764; 6, C. parapsilosis ATCC22019. Inhibition was tested independent of mannose binding on agar plates (Fig. 15). When grown in broth, the MBL+ lactobaciUi captured the yeast by mannan binding, and the killing rate was 100% against all six tested Candida species. The secreted anti-fungal agent is under study.

Claims

CLAIMSWe claim:

1. A method for altering the bacterial composition of the digestive tract or urogenital tract of a mammalian subject comprising administering to the subject bacteria at least one MBL-positive Lactobacillus strain.

2. The method of claim 1 , wherein at least two MBL-positive Lactobacillus strains are administered.

3. The method of claim 2, wherein the at least two MBL-positive Lactobacillus strains comprises a first MBL-positive Lactobacillus strain comprising a surface-bound MBL and a second MBL-positive Lactobacillus strain that secretes an MBL.

4. The method of claim 1 , wherein the Lactobacillus strain is an oral isolate characterized by the ability to co-aggregate with Saccharomyces cerevisiae and to bind to gpl20.

5. The method of claim 4, wherein the Lactobacillus strain is further characterized by the ability to bind to at least one HIV target cell.

6. The method of claim 4, wherein the Lactobacillus strain is further characterized by the ability to block HIV infection in vitro.

7. The method of claim 1, wherein the MBL-positive Lactobacillus strain is grown or derived from a strain deposited as PTA-5849 or PTA-5850.

8. The method of claim 1 , wherein the Lactobacillus strain is an vaginal isolate characterized by the ability to co-aggregate with Saccharomyces cerevisiae and to bind to gpl20.

9. The method of claim 8, wherein the Lactobacillus strain is further characterized by the ability to bind to at least one HIV target cell.

10. The method of claim 8, wherein the Lactobacillus strain is further characterized by the ability to block HIV infection.

11. The method of claim 8, wherein the Lactobacillus strain is further characterized by the ability to inhibit growth of at least one of Esherichia coli ATCC 14243, Staphylococcus aureus ATCC25923, Pseudomonas aeruginosa ATCC27853, Gardnerrella vaginalis ATCC 14018, Gardnerrella vaginalis ATCC49145, Mobiluncus curtisii subsp curtisii ATCC35241, Candida albicans ATCC 10231, C. krusei ATCC 14242, C. glabrata ATCC66032, C. tropicalis ATCC 13803, C. kefir ATCC46764, and C. parapsilosis ATCC22019.

12. The method of claim 1 , wherein the Lactobacillus strain is delivered to the subject delivered by mouth.

13. The method of claim 1, wherein the Lactobacillus strain is delivered to the subject vaginally.

14. The method of claim 1, wherein the Lactobacillus strain is delivered as a food product or a pharmaceutical composition.

15. A composition comprising MBL-positive Lactobacillus bacteria, the bacteria characterized by the ability to co-aggregate with Saccharomyces cerevisiae and to bind to gpl20.

16. The composition of claim 15, wherein the bacteria are further characterized by the ability to bind to at least one HIV target cell.

17. The composition of claim 15, wherein the bacteria are further characterized by the ability to block HIV infection.

18. The composition of claim 15, wherein the bacteria are an oral or vaginal isolate.

19. The composition of claim 18, wherein the Lactobacillus strain is grown or derived from a strain deposited as PTA-5849 or PTA-5850.

20. The composition of claim 15, wherein the Lactobacillus strain is lyophilized.

21. The composition of claim 15, wherein the composition further comprises a prebiotic.

22. The composition of claim 21, wherein the prebiotic is selected from the group consisting of glucono-delton-lactone and mannitol.

23. A method of evaluating a Lactobacillus isolate for its suitability for use in the method of claim 1 or the manufacture of a composition of claim 15 comprising testing the isolate for the ability to co-aggregate with an enveloped virus or a yeast comprising mannose its cell wall, for the ability to bind a mannose containing viral coat protein, for the ability to bind at least one HIV-target cell, for the ability to block HIV infection, or for the ability to inhibit the growth of a Candida species or of a bacterial vaginosis associated bacterium.