CA2165951A1 - Treatment of diabetes using phosphorylated insulin - Google Patents

Abstract

Phosphorylated insulins, which can be prepared from chemically extracted pharmacological insulins by gentle treatment with phosphorous oxychloride, have been shown to have reduced bioactiviy by mouse convulsion assay, but such phosphorylated insulins reduce hyperglycemia when administered to diabetic subjects without inducing hypoglycemia.

Description

~ 2 1 6595 1 9~/0054g . ~ ~ ~C~ ~100269 ~r,P~tment of ni~h~te~ inrt phosDhorylated TnC~ n F~ of TnYent~on This invention relates to the tre~tment of ~Ahetes ~nd no~el ~orms of i~ttl ~n therQfor.

R~Ck~LO~ of T~ven~on and prior ~rt Since $ts first extraction by Banting and Best in 1921, insulin ha~ been admini~tered to countle~s thousands of dia~etic p~tient~ w~th dramatic and life-~aving eff~cts Not ~uL~Lisingly~ ~n~ has been intensi~ly ~tudied over the ~ucceeding ~eventy years as a model peptide and elucidation of its cry~tallographic structure, peptide ~equence, radio ligand as~ay ~nd chemical synthesis reprcsent ma~or ach~evements in the h~tory of molecul~r biology Despite all of the work which ha~ b-en done, the original method ~or self administr~tion of inculin, namely su~cutaneous in~ection, per~$sts to thi~ day and it is widely ~umed it is this non-phy~iological route that result~ in the well known very poor control of blood glucose It is also well known that when extracted in~ulins are in~ected hypoglycemia is readily produced by a ~mall overdose, whereas when natural insulin delivery i8 restored to a diabetic patient, such as by pancreatic or I~let of Lang~rhans cell tran~plantation, hypoglycemi~ is not a problem It ~ppears, therefore, th~t while extracted insulins are very si~ilar to natur~l insulins, tbey are not, -SUBSTITUTE SHEET

WOg5/00549 ..
. s,. : -. PCT/CA93/00269--~ ~ ~ 5 ~ c ~ ~ J ~
in fact, the same. Indeed, over the years, the st~ndard ~ssay (mouse convulsion ~ssay) to determine the potency of ~n extracted (pharmacological) insulin measures the amount of insulin necessary to induce rapid hypoglycemia in a typical normal ~nim~l where~s hypoglyce~a ic not ~Gduced at ~ll with native (indogenous) insuL By this ~ssay, therefore, endogenous insulin coul~ be said to ~e non-potent.
The present inventor hypothesizes that the chemical ex~raction process commonly employed to Qb~in insulin from recombinant sources or from bovine or porcine pancreas de-natures or degrades the native (indogenous) insulin or th~t ~he handling of the pharmacological insulin in some measure degrades the material. It is interesting to note that Banting and ~est, in 1921, reported one water solubilized extract which normalized blood sugar in a diabetic dog but did not produce hypoglycemia and which was subsequently abandoned as unproductive, in favour of more vigorous chemical extractions with a more potent hypoglycemic action.
Apparently Banting and 8est did not appreciate the significance of their results which indicate to the present inventor that some subtle degradation of indogenous insulin occurs during chemical extraction in normal ~nimals.
Following major studies into the human biochemical fuel cycle and many of the constituent substrate-hormonal interdependencies using a comprehensive metabolic simulator SUBSTITUTE SHEET

5 t .which. included detAils of the energy cycles of carbo-hydrate-, protein-, and fat- derived fuel substrates, the inventor herein postul~tes that the chemical production of phar~acological insulin strips phosphate groups from the outer surfaces o~ thè ~nc~ molecule from residues widely known to be ph~ ~ orylatable and that pho~phoryl~tion of chemically extracted insulin is cap~ble of producing an insulin which (a) will not invoke hypoglycemia in normal subjects (b) will reduce hyperglycemia in ~iAhetic sub~ects but only to normal levels and (c) is not dose-response deren~Dnt.
Phosphoryl~tion of amino acids, proteins and peptides inclu.ding insulin is not novel. Attention i directed to U.S. Pstent 4,705,845 issued 10 November 1987, but such modi~ied insulins retain about 85% of the bio-activity of the unphosphorylated starting material and hence induce hypoglycemia in diabetic and non-diabetic subjects. Clearly such phosphoryl~ted materials contain a considerable amount of unreacted or de-graded materi~l.

Ob~ect of the Invention Thus, it is an object of the present invention to provide a novel, su~stantially pure phosphorylated product which will not invoke hypoglycemia in normal subjects and will reduce hyperglycemia in diabetic ~ubjects to normal levels.

SUBSTITUTE SHEET

~ 21 6 5~ $ ~ PCT/CA93/00269 Another object of the present invention is to provide a ~o~ass for making novel insulins.

Rr~ef S~te~ent of Tnvention By one aspect of the prese~t lnvention there is provided a ~ubstantially pure phD~phorylated insulin, for use in the treatment of dia~etes, which reduces hyperglycemia without inducing hypoglycemia when ~ri n~ ~tered to A diabetic subject.
By another aspect of this invention there is provided a method for producing a phosphorylated insulin which reduces hyperglycemia without inducing hypoglycemia when A~m;~;~tered to a diabetic subject, said method comprising:
~) dissolving a selected extracted in~ulin in a mixture of dry dimethyl formamide and concentrated phosphoric acid;
(b) adding phosphorus trichloride oxide and agitating s~id mixture at a temperature below about 10C;
(c) ad~usting pH to about 7.4;
(d) centrifuging said solution and recovering a ~upe2,.ate;
(e) dialyzing said supernate against ammonium chloride 80 as to remove salt;
(f) ~eparating by iso-electric focussing;
(g) purifying by ion e~h~ge chromatogr~phy;
(h) isolating a phosphorylated insulin product by lyophilization.

SUBSl ITUTE SHEET

2 ~` 6 ~ 9 ~
~ 3 ~ ~CT/CA93/00269 By yet another aspect of this invention there is provided a method of treating diabetes mellitus without ~nducing hypoglyce~ comprising administering to a diabetic ~ubject an effect~'~e ~mount of a ~ubstantially pure phosphorylated insulin so ~s to reduce hyperglycemia and maintain normoalycemia.

Rr~ ef Descr~tion of Draw;nas Fig. 1 is a ch~rt showing the amino acid sequences of various mammalian insulins;
Fig. 2 is a schematic sketch of the amino acid structure of porcine insulin;
Fig. 3 is a graph showing plasma insulin after injection of regular and phosphorylated insulin, against time;
Fig. 4 is a graph showing plasma glucose after injection of regular and phosphorylated insulin against time; and Fig. 5 is a graph illustrating plasma glucose levels with time after injection of different doses of phosphorylated insulin.

Detalled Descri~tion of Preferred ~mbodim~nts The amino acid sequences for the A and B chains of a number of mammalian insulins are shown in Figure 1. Given that phosphate groups may only be substituted on Ser, Thr and Tyr residues, it can be seen that mammalian insulins SUBSTITUTE SHEEl`

-WO 9~100549 ~ 5 ~
21 ~9 PCT/CA93/00269--h~ve between 8 and 11 residues where phosphate groups may be ~ubstituted. Serine re~idues ~t Bs And A12 are conserved, except that fiubstitution of the threonine at A12 in Guinea Pig is equivalent to a fierine ~eSidue ~t that position insof~r ~s phosphorylation is concerned. For porcine insulin, all the possible bin~in~ sites ~re illustrated in ~igure 2. Structurally, those at A8, A9, A12 ~nd 89 are ~patially most geometrically defined becAuse of their proximity to the inter- and intra-molecul~r disulphide ~ridges as shown.
Energy minimization calculation~ indic~te no ma~or changes from the starting configuration of porcine insulin when substituted at all three serine residues. Root mean square displacements from the initial conformation are detailed in Table 1.
Tabl- ~
Phosphorylation of Insulin: Root Mean Square Displacements from Initial Conformation An~strom Units All atoms 1.421 Baclchon~ 1. 019 Residue 89 0.713 Residue A9 0.294 Residue ~12 0.340 The specific locations of the three serine residues (A9, A12, B9) in the A and B chains clearly place the added SlJBSTiTUTE S~3EE~

-vo 95/00549 phosphate y~OUpS at the exposed surfaces of the prominent loops formed by the two overlapping (inter- and intra-chain) disulphide bridges (Fiqure 2). This raises the likelihood that other structuresj~w~ch may be ~imilar to mammali~n insulins but which may ha~e other biocompatible polymer or aminc ~cid backbones and whic~ may or may not be peptides, may be synthesized according to coordinates derived from energy minimization lrinciples, such th~t one or more phosphate groups are precisely located on the surface of the molecule and which would have similar bioactivity. Such compounds could be made orally acceptable using known t~rhniques.
It i8 known that polar amino acid residues with aliphatic or aromatic hydroxyl ~ou~ can be rhosrhQrylated both enzymatically and chemically. Using the chemical route it is somewhat easier to separate and purify the resulting crude material, and consequently this is the preferred route. Early workers attempting to phosphorylate insulin reported a product with 38% o~ the biological activity of the original, using the mouse convulsion assay, but apparently did not test it in dia~etic animals. The ~nvestigators attributed the reduced bioactivity to damage to the peptide chain. In U.S. Patent 4,705,845 a phosphorylation process is described in which insulin was dissolved in an organic solvent and reacted at reduced temperature with concentrated phosphoric acid to yield a SllJBSrlTU~ S~EEI~

.f .~ ,~. ? ;~.

~ ~ 6~ 9 51 PCT/CA93/00269 product havinq 85~ of the bioactivity of the unmodified insulin. Thi~ minimized supposed damage to the peptide chain and r~ e~ the ~phQsphorylation which occur~ ~t low pH.

mDle 1 Method for makins ~hos~horylated ~nSultn 20 mg of porcine insulin (Connaught-Novo, ~G~,.Lo) was dissolved in ~ 4C mixture of 9S0 ~1 dry dimethyl-formamide (DMF) and 50 ~1 of concentrated phosphoric acid which was prepared by heating 85~ aqueous rhosphoric acid at 160C for ~ hours. 5 ~1 of POCl~ (phosphorous trichloride oxide) dissolved in 15 ~1 DMF was added and the mixture was ~haken at 4C overnight. Approximately 1 ml cracked ice was added ~nd the pH adjusted to 7.4 with lON NaOH. The ~olution was brought to 5 ml and centrifuged. The ~upernate was extensively dialyzed against 50 mM ammonium bicarbonate pH
7.5 to remove salt and then separated by i~o-electric focu~ing. Samples were purified by ion-exchange chromatography and isol~ted after gel filtration by lyophilization from the ammonium bicarbonate solution.
Insulin conc~"L,~tion was determined by immuno-assay. Only separated samples with multiple (3 or 5) negatively charged adducts were used. Characteristically the modified product was intact, monomeric in~ulin with about 5-10% being unmodified, according to the chromatographs. The crude products were readily soluble in water at neutral pH. In SUBSTITUTE S~tEET

~tl:~5~ i VO 95/00549 = ~ ~ ? ~ ar~ s order to prolong its biological action ~ , ~otamine was added to complex with the phosphorylated ;~C~ n and thereby to reduce its rate of vascular entry ~ecause of aggregate cize .
~ he phosphorylated insulin produced in Example 1 was extensively tested in laboratory dogs. The dog is an excellent model for human metaboli~m. Pancreatectomy makes it metabolically unstable, with difficult to regulate glycemia while critically dependent on daily exogenous insulin replacement. Glycemia i~ very difficult to stabilize when co~ventional methods are used and even when extracted insul:~.n is pumped intravenoucly and continuously all according co open-loop control methods. Small changes of a few percent in the delivery rates of extracted insulin have major effects on fasting giycemia. Although normal blood glucose levels can be achieved in this model using a closed-loop artificial endocrine pancre~s instrument, the decline of biood glucose into hypoglycemia almost always has to be averted by co-infusions of dextrose or glucagon, even if pancreatic or extra-pancreatic glucagon is present. With careful selection of the parameterC in the specific algorithms, normoglycemia can be achieved and maintained, but this is not robust. Any over-insulinizing of the subject immediately provokes the need for such counter-regulatory co-infusions.

SU~3STIl-IJTE SHEET

t ~ r W0 95/00549 Z 1 ~; `3 9 5 ~ '? ~ ~
. ' PCT/CA93/00269 ~mple 2 ~nimal Studies Nine dogs of initially normal body weight were studied in accordance with Instit~ ~ on~l Guidelines for animal experiments. Exteriorized, indwelling catheters were placed for intra-portal insulin administration and peripheral venous blood sampling. All were ~tudied initi~lly as normal~. C~ quently all were pancreatectomized; five were auto~rafted and the remaining four were treated pharmacologically both with the modified insulins described above in Example 1 ~nd the correGponding unmodified conventional extracted insulins. Small venous blood samples were drawn via the indwelling blood s~mpling catheter at 30-60 minute intervals for plasma glucose determination in the subcutaneous experiments and at -10,0,3,6,10,20,30,60 minutes in the intravenous experiments. Intraportal dosages of regular porcine insulin unless otherwise specified were standardized at 210 pMol/kg ro. 03U/kg) body weight and 200+40 pMol/kg of phosphorylated insulin. All animals were conscious and ambulatory during the studies.
At the conclusion of experimental protocols, it was verified that all nine of the animals were diabetic by extirpAting the grafts (from the transplant recipients) or withdrawing insulin therapy (from the pancreatectomized animals). In ~ll cases, rapidly evolving hyperglycemia became life-threatening. For survival, insulin treatment had to be re-initiated.

SUBSTITUTE SHEET

O9st~49 ~ 93/00269 Plasma glucose was determi~e~ using a glucose analyzer (Beckman Instruments, Fullerton, CA). IRI was assayed using porcine insulin standard an~ 125I porcine insulin tracer (Novo Research Institute, Gentofte, Denmark), anti~erum and a dextran-coated charcoal ~epar~tion technigue (Albi~ser et ~1, Di~betes 35:97-100, 1986). In the studies using porcine phosphorylated insulin, the ~bove assay was performed as for porcine insulin, but using equimolar ~tandards of phosphoryl~ted insul in. Parallel standard curves were obtained. Pl sma samples when diluted gave values that fell along the phosphorylated insulin st~n~rd curve. All values were expressed afi mean ~ SEM unless otherwise indicated.

Tntravenous Studies Figures 3 and 4 demonstrate that in the normal dog, intra-portal administration of phosphorylated insulin has a minor effect on blood glucose (~ig. 4), despite the expected increases in measured plasma insulin concentration (Fig. 3).
At ~5 minutes this minor drop of 0.2 mM reflects the presence of the small y~po,Lion (5-10%) of unphosphorylated insul in in the crude preparations used in these studies.
Control studies with simil~r amounts of purified porcine regular insulin showed the expected insulinemia and hypoglycemia characteristic of thi~ extract and route of administration. Yet, even a 2-fold larger dose of phosphorylated insulin is without ma~or hypoglycemic effect.

SUBSTITUTE SHEE r :

. PCT/CA93/00269 Sub~u~Aneous Studies Subsequent to p~ncreAtectomy the animals were ~tudied repeatedly, now in the post-~bsorptive, hyperglycemic state of 14-20mM, because the conventional insulin depot admini~tered the previous day had become depleted. These ~xperiments indicated that over an approximately 5-fold r nge of 30-150 nMol, phosphorylated ~n~ was adeguate to lower ~lood glucose concentrations and to m~ intain euglycemia for 2-6 hours in the fasted animal. As shown in Figure 5, when crudely phosphorylated insulin was injected subcutaneously at time zero (about O900h), small specific doses of approximately 4 nMol/kg lowered pla~ma glucose to the normal range of 4.5-6.0 mM. When 4-fold larger doses .were given, the effect occurred more rapidly and lasted longer. Early hypoglycemia never occurred, but there was a delayed tendency to levels beneath the normal range with greater than 4-fold larger doses. Similar relative effects occurred consistently regardless of the starting plasma glucose concentration. It is important to recognize that a 4-fold larger dose of regular (unphosphorylated) insulin could be fatal to the dog.
Repeating the smaller dosage injections of phosphorylated insulin at 6-8h intervals (not ~hown) served to maintain pla~ma glucose in the range normal for post-a~sorptive dogs (shown here as the stippled area). An~mal~
treated with twice daily injections using split dosages of SUBSTITUTE SI~EET

~2 ~6 5 S``5 ~ ?~,c~
~ " J~ PC ~3/00269 regul~r phosphorylated insulin and protamine complexed phosphorylated insulin were easily m~intained before And after meals within the normal range of glycemia (4.5-6.0mM) for extended periods, up to 30 dAys. With continued treatment, all animals dev~ope~ ralrAhle fat-like deposits in th~ areas of repeated su~cutaneou~ injections with phosphorylated insulins. Such li~o ~..esis WAs not observed in ~nimals similarly in~ected for periods of 1-3 years but with purified porcine co.~ve..~ional insulins .
In contrast to the marginal stability achievable with the injection or infusion of extracted insulins, auto-transplantation of the ~n;~e pancre~s immediately lev~l~ed the diabetes. In the experiments herein it robustly normaYized the blood glucose conce~trations and ameliorated the metabolic states of the five recipients even though the route of administration was peripheral rather than portal and the amount of pancreas grafted was variable.
Hypoglycemia never occurred. The results were similar to those obtained with phosphorylated insulin simply injected subcutaneously.
The liver is known to extract about half of the insulin presented to it on the first pass. Apparently phosphorylated ;nClll in is not extracted to the same extent.
This results in the higher peripherAl concentrations at the peak and the slower decline to fasting levels observed with phosphorylated insulin (Fig. 4). Despite the almost two-SUBSTITUTE SHEE r L
~ ~ ~ 5 9 51 PCT/CA93/00269 -fold higher phosphorylated insulin concentrations in the periphery, pl~sma glucose levels Are essentially unaffected.
The prompt and powerful plasma glucose lowering effect of regular, unphosphorylated insu~ is clearly demonstr~ted.
In choosing the amount o~ ~egular insulin to be used, attempts are made to keep the nadir in glucose levels preciGely within the physiological r~nge. However as phosphorylated insulin does not provoke the same drop in plasma glucose levels, on one occasion a double dosage was administered (solid line Fig. 4). This provokes a small decre~se in plasma glucose (~bout 0.5mM) that i~ consistent with the effect~ of a small fraction of unphosphorylated insulin in the phosphorylated material.
Extracted insulins do not robustly stabilize diabetes, when delivered by intensive conventional methods (including multiple daily subcutaneous injections or continuou~
intravenous infusions), but only when delivered intravenously by ~ closed-loop feedbac~ system responding to minute-by-min~te ~lood glucose measurements. However phosphorylated insulins do robustly stabilize diabetes, even when widely dif~erent amounts are administered subcutaneously and independently of any closed-loop blood glucose cG..~lol system. In the animal experi2ents le~G~Led herein, phosphorylated insulin exhibits effects entirely conqistent with all of the predictions made for it.
Experimentally, even crude chemical phosphorylation of SUBSTITUTE SHEET

~,VO 95/0054g ; ~ S 9 5 1 extrActed insulin restored to it many i~no~Sall of the hypothetically ideal biological characteristics of natural insulin: (i) it had no immediate hypoglycemic effect when blood glucose levels we~e ~ n the normal range, only when gluco~e levels were well ab,ove normal; (ii) it enabled the metabolic regulatory merh~n~m to proceed and stable normal glyce~ia resulted almost independently of dosing, even when administered subcutaneously; (iii) it certainly had (macroscopic) anti-lipolytic effects, as evidenced by the lipogenesis in the injection areas. Parentetically, these animals also showed a tendency to regain body weight (1-3kg) ,lost following pancreatectomy - a process unachievable with on-going management using extracted insulins.

SUBSrlTL:TE SHEET
-

Claims (12)

Claims

1. A substantially pure phosphorylated insulin, for use in the treatment of diabetes, which reduces hyperglycemia without inducing hypoglycemia when administered to a diabetic subject.

2. A substantially pure phosphorylated insulin as claimed in claim 1 containing at least three phosphate groups.

3. A substantially pure phosphorylated insulin as claimed in claim 2 wherein three of said phosphate groups are substituted at serine residues in said insulin.

4. A substantially pure phosphorylated insulin as claimed in claim 3 additionally phosphorylated at threonine residues in said insulin.

5. A substantially pure phosphorylated insulin as claimed in claim 4 additionally phosphorylated at tyrosine residues in said insulin.

6. A method for producing a phosphorylated insulin which reduces hyperglycemia without inducing hypoglycemia when administered to a diabetic subject, said method comprising:
(a) dissolving a selected extracted insulin in a mixture of dry dimethyl formamide and concentrated phosphoric acid;

(b) adding phosphorus trichloride oxide and agitating said mixture at a temperature below about 10°C;
(c) adjusting pH to about 7.4;
(d) centrifuging said solution and recovering a supernate;
(e) dialyzing said supernate against ammonium chloride so as to remove salt;
(f) separating by iso-electric focussing;
(g) purifying by ion exchange chromatography;
(h) isolating a phosphorylated insulin product by lyophilization.

7. A method of treating diabetes mellitus without inducing hypoglycemia comprising administering to a diabetic subject an effective amount of a substantially pure phosphorylated insulin so as to reduce hyperglycemia and to maintain normoglycemia.

8. A method as claimed in claim 7 wherein said phosphorylated insulin is administered by subcutaneous injection.

9. A compound capable of reducing hyperglycemia without inducing hypoglycemia when administered to a diabetic subject, said compound having a chemical structure such that at least one phosphate group is located at a selected location on the surface thereof as defined by energy minimization computations of known molecular structures.

10. A compound as claimed in claim 9 wherein said structure comprises groups selected from amino acids, biocompatible polymers and mixtures thereof.

11. A compound as claimed in claim 10 wherein a plurality of phosphate groups are located on the surface of said structure.

12. A compound as claimed in claim 11 wherein said compound is orally acceptable.