WO1998026804A1

WO1998026804A1 - Bioassay for compounds that prevent diabetic complications and agents identified thereby

Info

Publication number: WO1998026804A1
Application number: PCT/US1997/023281
Authority: WO
Inventors: Kim Fechtel; Ying Peng; Peter G. Smith
Original assignee: University Of Kansas Medical Center
Priority date: 1996-12-17
Filing date: 1997-12-16
Publication date: 1998-06-25

Abstract

The present invention provides an in vitro and in vivo bioassay utilizing mammalian tissues such that the action of agents that are effective in blunting the impact of hyperglycemia can be detected. The assay can be completed in approximately 10 days, is technically simple, has exquisite sensitivity and produces a graded response such that therapeutic agents can be identified and their relative efficacies can be assessed. Compounds exhibiting a positive result in this bioassay have the potential of being preventative agents in the following disease processes: diabetes (type I and type II), atherosclerosis, birth defects, and the recurrence of certain cancers. In employing the in vitro bioassay of the present invention, a class of compounds, folates, folate derivatives and pterins, has been identified that prevents the measurable biological consequences of hyperglycemia in a mammalian model. This novel class of compounds has not been linked previously to diabetes. Folates, folate derivatives and pterins have very low toxicity and make excellent preventative therapeutics.

Description

BIOASSAY FOR COMPOUNDS THAT

PREVENT DIABETIC COMPLICATIONS AND

AGENTS IDENTIFIED THEREBY

Federal Funding Legend

The present invention was supported in part by Federal funds. The United States Government may have rights to this invention.

BACKGROUND OF THE INVENTION

Field of the Invention

The present invention relates generally to a bioassay for identifying pharmacologically useful compounds. In particular, the present invention relates to a bioassay for identifying compounds that are effective in preventing diabetic complications resulting from hyperglycemia. Additionally, certain classes of compounds, folates, folate derivatives and pterins, have been identified by the bioassay as being effective in preventing neural tube defects resulting from hyperglycemia.

Description of the Related Art

Anencephaly and spina bifida are common birth defects leading either to perinatal death or to physical handicap and mental retardation in survivors (McLone, D.G. and T.P. Naidich, Myelomeningocele , Ch. 6, Disorders of the Developing Nervous System: Diagnosis and Treatment,, Hoffman and Epstein, eds., Blackwell Scientific Publications (1986)). These defects arise due to the failure of the neural tube to close during the earliest steps of neurulation. Anencephaly is attributed to the failure of the rostral neuralepithelium to close in the cranial neural tube and is represented in animal models as exencephaly (Padmanabhan, R. Ada. Anat. Basel 141 : 182-92 (1991 )). Exencephaly is visible following the completion of neural tube closure as a pair of open cranial neural folds. At term in anencephalic infants or animal models, the majority of the cranial neural elements are absent; a condition that is incompatible with life independent of the maternal environment.

Spina bifida arises through a similar process in which the neural tube fails to close resulting in an open lesion within the spinal column. The increased risk of these congenital defects in families in which a previous infant was born with neural tube defects (NTDs), suggests that genetic factors contribute to the etiology of these malformations; however, in the absence of definitive twin studies, it is difficult to separate whether familial factors are genetic or environmental (Campbell, et al., Teratology 34: 171 -87( 1986); and Ramos-Arroyo, M.A., Ada. Genet. Med. Bemellol-Roma. 40:337-44 (1991)).

The best evidence that genetic factors play a role in NTDs has been obtained from studies in the mouse in which mutations in single genes produce an elevated incidence of NTDs (see Gruneberg, H., Journal of Genetics 52:52-67 (1954); and Copp, et al., Progr. in Neurobiology 35:363-03 (1990); Gunther, et al., De ve lopment 120:31 19-30 ( 1994)). Moreover, the observation that targeted mutations in a number of murine genes produce exencephaly in a subset of homozygotes makes a strong case for genetic factors in the etiology of human neural tube defects, given the similarities in neural development among vertebrates Lufkin, et al., Cell 66: 1105-19 (1991); and Dolle, et al., Proc. Natl. Acad. Sci. USA 90:7666-70 (1993)). However, with the possible exception of openbrain (Gunther, et al., Development 120:31 19-30 ( 1994)), in all cases murine mutations in single genes fail to produce Mendelian proportions of affected animals, indicating that the pattern of inheritance remains best explained as multifactorial.

The contribution of environmental factors to the etiology of NTDs is well established. A number of teratological agents have been shown reproducibly to cause NTDs in both humans and animal models (Campbell, et al., Teratology 34: 171 - 87( 1986); and Copp, et al., Progr. in Neurobiology 35 :363-03 (1990)). However, the etiology of neural tube defects is poorly understood and no pathways have been defined clearly linking the application of a teratogen to the production of neural tube defects in treated embryos. The absence of a well-developed animal model has hampered progress in identifying the pathways leading to NTDs, as in most murine mutant strains the penetrance is low and affected embryos can not be distinguished from unaffected ones prior to the obvious manifestation of the delay in neuropore closure.

Clinical studies have shown that maternal insulin- dependent diabetes mellitus is one of the highest risk factors underlying neural tube defects in humans (see Becerra, et al., Pediatrics 85: 1 -9 ( 1990)). It has been found that maternal supplementation with the B vitamin folic acid prior to conception and throughout the first trimester of pregnancy prevented the majority of NTDs in infants born to women with a previous NTD pregnancy (high-risk) and also in infants born to an unselected European cohort of women planning a future pregnancy (see A.E. Czeizel and I. Dudas, N. Engl. J. Med. 327: 1832-35 ( 1992)), demonstrating that these congenital defects are preventable. However, the efficacy of folates in the prevention of NTDs has not been tested in diabetic women.

Accelerated atherosclerosis and NTDs are common complications of diabetes mellitus and both are likely to arise through hyperglycemia-induced metabolic disruption. Similar disruptions in gene expression and adhesion occur during the progression of both disease processes, suggesting that similar biochemical mechanisms are involved. Recent evidence suggests that disruptions in cellular oxidation/reduction (redox) homeostasis caused by hyperglycemia is likely to be the causal event underlying or accelerating both disease processes in diabetics. The development of a simple assay in which the prevention of mammalian hyperglycemia-induced NTDs could be assessed would provide a rapid avenue to the development of pharmaceuticals that prevent diabetic complications and are applicable to other chronic, progressive diseases as well.

The prior art is deficient in providing a sensitive, technically simple bioassay for identifying compounds that are effective in preventing neural tube defects (NTDs), and other complications resulting from hyperglycemia. Additionally, the prior art is deficient in providing effective compounds to prevent NTDs in infants of diabetic women. The present invention fulfills these long-standing needs and desires in the art. SUMMARY OF THE INVENTION

One object of the present invention is to provide a bioassay for identifying compounds that are effective in preventing diabetic complications resulting from hyperglycemia. In addition, these compounds may have a broad utility in preventing several chronic, progressive diseases.

In an embodiment of the present invention, there is provided an in vitro bioassay to determine agents which can blunt the deleterious effects of hyperglycemia in an animal or its offspring, comprising the steps of: selecting an agent to be tested; dissecting rat embryos from pregnant dams; culturing said dissected rat embryos in hyperglycemic rat serum and said agent to be tested for a pre-determined time; and examining said cultured rat embryos to determine if abnormalities exist, wherein if abnormalities exist, said agent does not blunt deleterious effects of hyperglycemia, and if abnormalities do not exist, said agent does blunt deleterious effects of hyperglycemia.

In one embodiment of the present invention, there is provided a method of identifying compounds that are effective in preventing diabetic complications resulting from hyperglycemia, consisting of the steps of dissecting of nine-and-one-half day rat embryos, culturing said dissected embryos in hyperglycemic rat serum for a pre-determined time such as 30 hours, and assessing said dissected embryos for morphological defects at the end of the culture period. Agents that exhibit efficacy in blunting the impact of hyperglycemia are likely candidates for preventing neural tube defects. Once an agent has been identified by the bioassay of the present invention, the efficacy of the agent can be evaluated further by determining the transcript accumulation profiles of the Krox-20 and the Hoxb-1 developmental control genes.

Yet another embodiment of the present invention provides a positive control whereby folates, pterins or folate derivatives are added to the culture medium one hour prior to the addition of glucose to the medium.

Another object of the present invention is to provide an in vivo bioassay to determine agents which can blunt the deleterious effects of hyperglycemia in an animal or its offspring, comprising the steps of: selecting an agent to be tested; inducing insulin-dependent diabetes in dams following a glucose challenge; treating said dams with insulin; establishing pregnancy in said dams; withdrawing insulin from said dams; administering glucose and said agent to be tested to said dams; dissecting embryos from said dams; and examining said cultured rat embryos to determine if abnormalities exist, wherein if abnormalities exist, said agent does not blunt deleterious effects of hyperglycemia, and if abnormalities do not exist, said agent does blunt deleterious effects of hyperglycemia.

Another object of the present invention is to provide classes of compounds effective in preventing diabetic complications resulting from hyperglycemia. This class of compounds, pterins, folates, or folate derivatives, was identified by the bioassay of the present invention.

Other and further aspects, features, and advantages of the present invention will be apparent from the following description of the presently preferred embodiments of the invention. These embodiments are given for the purpose of disclosure.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows that Krox-20 transcripts persist abnormally in rhombomere three in rat embryos cultured in hyperglycemic serum.

Panels a-f: Embryos cultured in euglycemic rat serum reproduce the dynamic profile of Krox-20 transcript accumulation, previously described in the mouse, in which the reduction of rhombomere three signals precedes the loss of signal in rhombomere five (Wilkinson, et al., Nature 337: 461-64 (1989); and Wilkinson, et al., Nature 341 :405-09 (1989)).

Panels g-I: Embryos cultured in hyperglycemic rat serum show an abnormal persistence of Krox-20 transcripts in rhombomere three. The embryos depicted in panel a, b and g, h were cultured 16 hours and had 9- 10 somite pairs; the embryos presented in panels c, d and e, f were cultured 30 hours and had 14 somite pairs (c, d and i,j ); or 15 somite pairs (e, f and k, 1). There is a loss of overt segmentation (panels i and k compared to panels c and e), abnormal adherence of the neuralepithelium within rhombomere 4 (panel i compared to panel c) and irregularities with prolapsed or convoluted regions within the neuralepithelium (panel k relative to panel e) in embryos cultured in hyperglycemic medium. Embryos cultured 30 hours in hyperglycemic media had exencephaly; embryos cultured in euglycemic media were morphologically normal. Left-side photographs are with bright field illumination (a, c, e and g, i, k); right-side photographs are with dark field illumination (b, d, f and h, j, 1); embryo culture procedures were followed as described herein in Example 1 , except that two embryos were cultured in 2 ml of immediately heat-inactivated rat serum. Coronal sections are magnified 120x; sagital, 60x. n: neuralepithelium; ot: otic pit; s: somites, h, heart; 3, 4, 5 : rhombomeres three, four, and five; ir: irregularities in the normally segmented neuralepithelium.

Figure 2. Hoxb- 1 transcripts fail to accumulate to expected levels during neural tube closure when embryos are cultured in hyperglycemic medium. Embryos cultured in euglycemic rat serum (panels a-f) exhibit the expected profile of Hoxb-1 transcript accumulation described for the mouse (Murphy, et al., Nature 341 : 156-59 ( 1989); Murphy, P., and R. Hill, Development 1 1 1 : 61 -74 ( 1991 )), with strong signal detected throughout the interval during which neural tube closure occurs. Embryos cultured in hyperglycemic rat serum (panels g-I) fail to accumulate expected levels of Hoxb-1 transcripts from the time of initiation site formation (panels g,h), to the point at which neural tube closure should be complete (panels k-1). Embryos were cultured for 16 hours and had 9-10 somite pairs (panels a, b and g,h), or for 30 hours and had 14 somite pairs (panel c,d and i,j); or for 30 hours and had 16 somite pairs (panel e,f and k,l). Preparation of embryos, in situ hybridization, photography and abbreviations as described in Figure 1 and Examples 1 and 2; nc: neural crest; white arrows indicate regions where Hoxb- 1 signal is expected.

Figure 3. Folinic acid prevention of hyperglycemia- induced exencephaly results in embryos bearing a closed neural lesion. Panel a: Sagital view of an explanted rat embryo cultured for 30 hours in euglycemic serum; note the well-formed otic pit (open arrow), optic vesicle (closed arrow) and clearly demarcated pharyngeal arches (*). Both forebrain (f) and midbrain (m) protrusions are apparent and the heart (h) is in the normal position, well below the second pharyngeal arch. The tail exhibits the expected curvature. Panels b and c: Sagital views of embryos cultured for 30 hours in hyperglycemic serum with 20(g/ml folinic acid added one hour prior to the addition of excess glucose. Panel b: In an embryo with a moderately malformed appearance, the optic vesicle is compressed (closed arrow), the otic pit is less well-formed and the pharyngeal arches have fused (single *). The tail has failed to rotate to the expected position. Panel c: In an embryo exhibiting gross malformations, the otic pit is undetectable, the pharyngeal arches are compressed to a single structure that is fused with the brain and the heart (single *). The majority of the expected forebrain mass is reduced and the midbrain is apparently absent. As in Panel b, the optic vesicle is compressed (closed arrow) and the tail has failed to rotate. Panels d, e and f: Panel d shows a dorsal view of (a) in which the well-formed, round otic pits (open arrows) are clearly in the same plane as the neuralepithelium. The hindbrain exhibits the expected open, unfused morphology. Black arrows indicate that the neuralepithelium is normally not adherent rostral and caudal to the otic pit. Panel e: shows a dorsal view of (b) in which malformations clearly are apparent. The otic pits (open arrows) are rotated out of the plane relative to the neuralepithelium and are less distinct. The neuralepithelium is abnormally adherent spanning from approximately rhombomere three caudal to the spinal cord (black arrows). A prolapsed region is apparent rostral to the fused neuralepithelium and the mes- metencephalic constriction is abnormally closely apposed (to the large black arrowheads). Overall, the hindbrain region is flattened. Panel f shows the dorsal view of the embryo in (c). In addition to the features noted in panel e, this embryo has no apparent otic pit and the neuralepithelium is twisted. Between the twisted regions, abnormally adherent sections are apparent (black arrow). Figure 4: Folate pretreatment is associated with an improvement in Krox-20 gene expression in hyperglycemic embryos; however adhesion and patterning defects are evident in embryos with a closed neural lesion. Panel a: Parasagital section an embryo with normal morphology. Note the clear hindbrain segmentation (numbers refer to specific rhombomeres) and the well-formed optic vesicles (large *). Panel b: Coronal section of a moderately malformed embryo revealing the uncharacteristic fusion of the hindbrain neuralepithelium (double arrowheads) and a prolapsed region encompassing rhombomeres 1 and 2 (below n). Visible segmentation is muted; rhombomeres are assigned based on the Krox-20 signal and the proximity to the malformed otic pit (open arrow). Panel c: Parasagital section of a grossly malformed embryo in which the forebrain in reduced, overt segmentation is absent and an extended hindbrain morphology is manifest and the otic pits are absent. Rhombomere assignments are based on Krox- 20 signal. Panel d: Darkfield image of the embryo in a indicates a slight delay in the disappearance of Krox-20 signal in rhombomere three in this 16 somite embryo. In embryos cultured in euglycemic medium, signal within rhombomere three is undetectable at this developmental stage. Panel e: Darkfield image of (b) reveals abnormally elevated levels of Krox-20 signal in rhombomere three in this 16 somite embryo. Panel f: Darkfield image of (c) reveals a Krox-20 molecular phenotype identical to that observed in exen c epha li c embryos, with a decrease in rhombomere five signal and enhanced rhombomere three signal, an abnormal molecular signature in this 14-somite embryo. Embryos were cultured in hyperglycemic medium for 30 hours with 20(g/ml folinic acid added one hour prior to the construction of hyperglycemic serum. In situ hybridization conducted as described in Example 4.

Figure 5. Folate pretreatment is associated with the recovery of Hoxb- 1 gene expression in hyperglycemic embryos. Panels a-c: Brightfield images of embryos that have been sectioned in the coronal plane reveal variable morphological irregularities within the hindbrain neuralepithelium including irregular segmentation (a), abnormal adhesion (b and c). Panels d-e : Darkfield images of the same sectioned embryos presented in panels a-c. Clearly visible is a moderate Hoxb- 1 signal in rhombomere 4 in these hyperglycemic embryos. Thus, folate pretreatment has normalized the gene expression profiles expected in rhombomere 4 which is absent in hyperglycemic embryos cultured in the absence of folate pretreatment (compare to Figure 2 panels g-1).

The appended drawings have been included herein so that the above-recited features, advantages and objects of the invention will become clear and can be understood in detail. These drawings form a part of the specification. It is to be noted, however, that the appended drawings illustrate preferred embodiments of the invention and should not be considered to limit the scope of the invention.

DETAILED DESCRIPTION OF THE INVENTION

It will be apparent to one skilled in the art that various substitutions and modifications may be made to the invention disclosed herein without departing from the scope and spirit of the invention.

As used herein, the terms "neural tube defects" or "NTDs" means a visible abnormal morphology including an open region of the neural tube that is normally closed at the developmental stage examined but can also include a closed neural lesion in which the neural tube is closed but exhibits abnormal morphology (for example : abnormally adherent, prolapsed, irregularly convoluted, loss of segmentation, etc.) and may also exhibit abnormal gene expression profiles (for example: Krox-20 or Hoxb- 1 or other genes typically differentially expressed within the neural tube). As used herein, the term "hyperglycemia" or "hyperglycemic" means having a abnormally high concentration of glucose. As used herein the term "euglycemic" means having a glucose level within the normal range for a non-diabetic animal. As used herein, the term "morphological defect" means the presence of an abnormal shape, composition, or conformation of an anatomical structure visible normally or under light or electron microscopy and distinct from that observed in normally developing animals. As used herein, the term "Krox-20" (also known as egr-2) means the gene or mRNA transcript encoding a transcription factor bearing a zinc- finger motif. Krox-20 transcripts accumulate in a dynamic temporal pattern restricted to rhombomeres three and five in the mouse (Wilkinson, D.G., et al., Nature 337: 461-64 (1989); and Wilkinson,et al., Nature 341 : 405-09 (1989)). As used herein, the term "Hoxb-1" means the gene or mRNA transcript encoding a transcription factor with a homeodomain motif and that contributes information to the matrix of developmental cues known as the Hox code. Hoxb-1 transcripts accumulate only in rhombomere four in the murine rostral neuralepithelium (Murphy, P., et al., Nature 341 : 156-59 ( 1989); and Murphy, P., and R. Hill, Development 111 :61-74 (1991)). As used herein, the term "pterin" refers to any member of the chemical family of substances known as pterins exhibiting a characteristic ring structure of the pterin moiety of folates. As used herein, the term "pterinaldehyde" means a chemical structure comprising a pterin ring with an aldehyde R group. As used herein, the term "pterinaldehyde synthetic product" means the product of the synthetic procedure of Thijssen (Anal. Biochem. 54: 609-61 1 ( 1973)), wherein pterinaldehyde is synthesized from folic acid using a hydrogen bromide solution and elevated temperature to effect folate cleavage. As used herein, the term "folate" refers to any chemical structure accepted commonly as a member of the folic acid family of chemical structures. As used herein, the term "folate derivatives" means any chemical synthetic product retaining specific features of folate structure, including but not limited to the pterin ring, or other components of folate structure. As used herein, the term "rhombomeres" means transiently segmented regions of the hindbrain known to comprise lineage restricted compartments (see Krumlauf, R., Trends in Genetics 9: 106-12 (1993); and Krumlauf, R., Cell 78: 191 -01 (1994), and references cited therein) . As used herein, the term

"neuralepithelium" means that region of the embryo that will give rise to the central and peripheral nervous system. As used herein, the term "exencephaly" means an open cranial neural tube evident at a developmental stage at which the neural tube is normally closed. Exencephaly is the accepted primary neural lesion leading to anencephaly postnatally. As used herein, the term

"anencephalic" refers to the condition where open cranial lesions are present. As used herein, the term "pharmacologically acceptable carrier" refers to any number of inactive chemicals that can be used to deliver the pterinaldehyde or folate derivative compounds to an individual to elicit a therapeutic response. A carrier is pharmacologically acceptable if its administration can be tolerated by the recipient human. Such a carrier, along with the therapeutic pterinaldehyde or folate derivative compounds of the present invention, is said to be administered in a "therapeutically effective amount" if the amount administered is physiologically significant. An agent is "physiologically significant" if its presence results in a change in the physiology of the recipient human. For example, in the treatment of disease, a combination of carrier and a pterinaldehyde compound which prevents or arrests further progress of NTDs in developing embryos would be considered both physiologically significant and therapeutically effective.

The present invention is directed to a bioassay to determine therapeutic agents that can blunt the impact of hyperglycemia, preventing the development or progression of complications of diabetes mellitus including, but not limited to, neural tube defects, congenital defects of the heart, caudal regression syndrome, atherosclerosis, kidney disease, neuropathy, retinopathy. The bioassay provides a simple, rapid, and useful screening tool for agents with the potential to blunt the impact of hyperglycemia. Specifically, the present invention is drawn to an in vitro bioassay to determine agents which can blunt the deleterious effects of hyperglycemia in an animal or its offspring, comprising the steps of: selecting an agent to be tested; dissecting rat embryos from pregnant dams; culturing said dissected rat embryos in hyperglycemic rat serum and said agent to be tested for a pre-determined time; and examining said cultured rat embryos to determine if abnormalities exist, wherein if abnormalities exist, said agent does not blunt deleterious effects of hyperglycemia, and if abnormalities do not exist, said agent does blunt deleterious effects of hyperglycemia.

Further, the present invention is drawn to an in vivo bioassay to determine agents which can blunt the deleterious effects of hyperglycemia in an animal or its offspring, comprising the steps of: selecting an agent to be tested; inducing insulin- dependent diabetes in dams following a glucose challenge; treating said dams with insulin; establishing pregnancy in said dams; withdrawing insulin from said dams; administering glucose and said agent to be tested to said dams; dissecting embryos from said dams; and examining said cultured rat embryos to determine if abnormalities exist, wherein if abnormalities exist, said agent does not blunt deleterious effects of hyperglycemia, and if abnormalities do not exist, said agent does blunt deleterious effects of hyperglycemia.

It is expected that the agents identified in the bioassay of the present invention may be employed as therapeutic substances for preventing or diminishing the progression of diabetic complications. In addition, the agents identified by the bioassay of the present invention are excellent candidates for treatment of other disease processes not associated with diabetes mellitus, but which also exhibit folate-sensitivity, such as cardiovascular disease and the progression of certain cancers. For example, the present invention provides classes of compounds, pterins, folates, or folate derivatives, effective in preventing diabetic complications resulting from hyperglycemia. This class of compounds was identified by the bioassay of the present invention. Further, these compounds are effective in preventing cardiovascular disease and neural tube defects.

For therapeutic applications, a person having ordinary skill in the art of pharmacology would be able to determine, without undue experimentation, the appropriate dosages and routes of administration of the pterinaldehydes or folate derivatives in the novel treatment for preventing diabetic complications resulting from hyperglycemia of the present invention.

The following examples are given for the purpose of illustrating various embodiments of the invention and are not meant to limit the present invention in any fashion:

EXAMPLE 1: Bioassay for Compounds that Prevent Complications Resulting from Hyperglycemia Clinical studies have shown that maternal insulin- dependent diabetes mellitus (IDDM) is one of the highest risk factors underlying neural tube defects in humans (Becerra, et al., Pediatrics 85 : 1-9 ( 1990)— the prevelance is especially high if infants with hydrocephalus are categorized as bearing NTDs. The in vitro culture of explanted rat embryos in hyperglycemic media approximates this condition in an experimentally manipulable model. The rat embryo model produces 65-80% exencephalic embryos following 30 hours of culture in hyperglycemic medium.

In vitro Bioassay

One embodiment of the in vitro bioassay of the present invention comprises the dissection of gestational day 9.5 rat embryos followed by culture of the dissected embryos in hyperglycemic rat serum for 30 hours. At 30 hours, the embryos are assessed for morphological defects. A test agent is added to the culture medium to assess whether the agent is effective in preventing neural tube defects in the cultured embryos. The efficacy of the tested agent is assessed by observing a set of morphological landmarks and determining the degree of restoration of normal development occurring in embryos.

Rat serum was prepared as a culture medium in the following manner: serum donor males (retired Sprague-Daly male breeders) were anesthetized with an injection containing 64.7 mg sodium pentobarbital and exsanguination was initiated when the male rat was unresponsive to a pressure applied to the tail. Withdrawn blood was collected in sterile centrifuge tubes and immediately centrifuged for 20 minutes in a clinical centrifuge (rpm 3,400). The serum was transferred to another sterile centrifuge tube with a sterile pasteur pipet and placed into a 56°C water bath for 30 minutes for heat-inactivation. Serum was cooled at room temperature until ambient temperature was achieved, approximately half an hour. Sera from different rats were pooled and 10 ml aliquots were prepared in 14 ml centrifuge tubes. Serum was stored in a -20° C freezer until used. Prior to use, serum was removed from the freezer and allowed to warm to room temperature, and was then placed in a 38°C incubator to reach 38°C. Immediately prior to use, the serum was gassed with O 2 - C O 2 - N 2 (5:5:90) for at least 1 minute using a sterile pipet connected to the gas outflow. The remaining serum, which had been thawed, was never refrozen. Although in these studies serum was prepared in the laboratory, such serum is now commercially available (Equitech, Inc.).

Timed-pregnancies were induced in normal, Sprague- Daly rat dams and gestational day 9.5 embryos are explanted according to established dissection procedures (New, D.A.T et al., J. Reprod. Fert. 35: 135-138 (1973); New, D.A. T. Biol. Rev. 53 : 81- 122(1978); and Morriss, G.M. and D.A.T. New J. Embryol. exp. Morph. 54: 17-35.(1979)). Female dams were anesthetized (32 mg sodium pentobarbital, Anpro Pharmaceutical Co., Inc.), the abdomen was opened and the uterus is removed to a dissecting dish containing Hank's balanced saline. The uterus was opened and the embryos were removed by squeezing below each embryo with surgical forceps. The decidua was opened and the embryo was removed with its investing membranes intact. The embryo was removed retaining the ectoplacental cone, amniotic sac and the embryo as an intact unit, with the exception of the Reichardt's membrane, which was retracted and torn to facilitate diffusion of substances from the medium to the embryo. This procedure results in nearly 100% viability of dissected embryos at the end of the culture period. Embryos were dissected quickly in Hank's saline (Sigma Chemical Co., Inc.) and transferred to rat serum prewarmed to 38° C and pregassed immediately before use for several minutes with a mixture of 5% oxygen, 5% carbon dioxide and 90% nitrogen (Helget Co., Inc.). Females then received another identical injection of sodium pentobarbital constituting a lethal dose for euthanasia.

Cultures were placed in an incubator where culture tubes undergo continuous rotation. The agent to be tested for efficacy in the prevention of diabetic complications was added to the cultures prior to the addition of the embryos in an appropriate dosage and cultures were continued for the specified time. Littermate controls were established in which all aspects of the culture conditions were identical with the exception of the addition of the agent to be tested. At the end of the culture period, the embryos are removed, washed in Hank's saline and observed under a dissecting microscope to assess the degree of morphological abnormality present in the embryos using the appearance of the neural tube (closed or open, normal or abnormal morphology) and other anatomical structures (pharyngeal arches, otic and optic pits, tail flexure, relative positioning of hindbrain, midbrain, forebrain, and heart, abnormally adherent, convoluted, twisted or prolapsed neuralepithelium,etc). An overall assessment of normal, mildly, moderately or grossly abnormal was assigned to each embryo and recorded in laboratory documentation. Somite number was recorded for each embryo to record its overall developmental achievement which may differ from the expected temporal stage based on gestational age.

In vivo Bioassay

In another embodiment of the invention, and as a complement to the in vitro bioassay, an in vivo bioassay was developed. Insulin-dependent diabetes was induced in female dams injecting 70 mg/Kg streptozotocin (Sigma Chemical Co., Inc., according to the manufacturers directions) following a glucose challenge. Blood glucose levels were determined using a OneTouch (Lifescan, Johnson & Johnson Co., Inc.) personal glucose meter using the manufacturers reagent test strips. A blood glucose reading over 300 one or two days after streptozotocin treatment was taken as evidence of manifest diabetes given the normal range of 40-65 found in non-diabetic siblings. Intensive insulin therapy was established using diluted Humulin Ultralente (Eli Lily, Inc., 1 :6) using the manufacturer's diluting solution (Eli Lily, Inc.). Three insulin injections were administered per day at roughly five hour intervals (8 am, 3 pm and 8 pm) seven days per week. For the first week, glucose levels were monitored prior to injection, subsequently a standard dose was established for each rat (3 insulin units of the Humulin diluted 1 :6 using the Microfine IV insulin syringes in 200 gram female rats). Insulin injections were delivered sub-cutaneously in the skin on the flanks of the hindlimbs . Timed-pregnancies were establi shed following overnight housing with a breeder male. On the morning sperm were detected in a vaginal smear, glucose testing was resumed a minimum of three times daily and normoglycemia was undertaken with intensive insulin therapy adjusting insulin regimens on an individual basis (1, 2 or 3 units of diluted Humulin Ultralente administered on a sliding scale with the lower dose for a glucose reading of 100-200, the middle dose for a glucose reading of 200-300 and the highest dose for readings over 300). On the evening of gestational day 8, insulin therapy was withdrawn and pregnant dams were given Glutose brand glucose gel (1-3 ml) five times per day to maintain hyperglycemia as well as exchanging the water for a 5% glucose solution. Thus, hyperglycemia was maintained for the comparable 30 hour period during which neural tube closure occurs in an in vivo model. Embryos were dissected on gestational day 11 and 100% exhibited open neural tube defects and were comparable to the lesions observed in vitro.

Agents to be tested in the in vivo bioassay are delivered orally, by inhalation, or by injection to the pregnant dam providing a whole animal model which is of substantial utility for a second stage bioassay or further drug development. Thus, this newly developed in vivo bioassay permits the extension and validation of in vitro findings and the determination of appropriate dosage schedules of therapeutic agents on a mg/Kg basis prior to their testing in human subjects.

All aspects of animal use were in compliance with federal regulations for animal use in experimental biomedical research and all required institutional guidelines were fully executed throughout the project.

EXAMPLE 2: Transcript Accumulation Profiles of Krox-20 and Hoxb-1

Further assessment of the effectiveness of an agent in preventing neural tube defects in embryos was made by evaluating molecular landmarks utilizing the in vitro bioassay. The transcript accumulation profiles of the Krox-20 and Hoxb-1 developmental control genes are abnormal in rat embryos cultured in hyperglycemic serum. The abnormal transcript profiles correlate with the failure of neural tube closure and provide independent developmental markers.

Tissue Preparation: Cultured embryos were fixed in

4% freshly prepared paraformaldehyde in IX PBS (phosphate- buffered saline: 13.7mM NaCl, 2.7mM KCl, 2.3mM Na₂HP0₄, pH

7.4) at 4° C overnight. Embryos were rinsed in IX PBS for 5 minutes, 3 times each and were then immersed in 0.5M sucrose in IX PBX overnight. The next day, embryos were embedded in tissue embedding medium (TBS, Baxter, Inc.) and maintained on dry ice until the TBS compound solidified. Specimens were put into a cryostat for 10 minutes to reach the same temperature as the prechilled cryostat (approximately -22° C). Cryosections of 9 μM thickness were taken and placed on slides that had been acid washed and subbed with Poly-L-Lysine as described above. These slides were stored in a box with desiccator in a -70° C freezer for up to 1 month.

The slides were prepared for hybridization first by fixation for 20 minutes in freshly prepared 4% paraformaldehyde in a jar on ice water, and were then rinsed in IX PBS twice, 5 minutes each. The slides were then treated with 2.5% acetic anhydride in 0.1M triethanolamine buffer, pH 8.0, for 10 minutes at room temperature, washed twice in 2X SSC (Freshly diluted from a 20X stock) for 5 minutes each, and dehydrated in a graded alcohol series (30%, 50%, 70%, 80%, 90%, 100%). The slides were allowed to air dry before hybridization.

In situ hybridization was performed employing frozen tissue that was sectioned either in the parasagital or coronal plane. Hybridization utilized single-strand RNA probes labeled with 3~> S - uridine-triphosphate prepared from either the murine Krox-20 clone (750 nucleotide Pstl/Apal fragment) or the murine Hoxb- 1 clone (800 base pair EcoRI fragment). In each case, 4 x 10" cpm/slide was used and hybridization was at 45° C for four hours; the processing followed standard procedures . Briefly, prehybridization was performed by immersion of the slides into a solution of 5mM MgCl2 (in IX PBS) for 10 minutes at room temperature, followed by a ten minute treatment in 0.25M Tris, pH 8.0, 0.1M glycine at room temperature. Hybridization solution consisted of 50% formamide, 2X SET (0.3M NaCl, 0.004M EDTA, 0.06M Tris-HCl, pH 8.0), 8% Dextran sulfate, 5X Denhardt's and 250 μg/ml yeast tRNA. The -^ ^ S-labeled riboprobe was diluted in an appropriate amount of hybridization solution and the mixture was heated to 60° C in a tightly capped centrifuge tube. An aliquot of 100(1 of hybridization solution was placed over the tissue sections on each slide and the tissues were covered with a clean, heat- treated glass coverslip (previously baked at 400° C overnight). The slides were then incubated in humidified chambers (sealed petri dishes containing water-soaked filter paper) at 45° C for 4 hours. Following hybridization, slides were washed twice, briefly, in 4X SSC to remove the coverslips and the majority of the unhybridized probe. The slides were washed with 2X SET, 0.1 % Beta-mercaptoethanol at 60° C for 15 minutes, followed by 2 rinses in 4X SSC. The slides were incubated in 40μg/ml RNase A in 3X SET at 37° C for 30 minutes and washed in a large (4 liter) volume of IX SSC at room temperature by suspending a rack in a 4 liter beaker and gently stirring with a magnetic stir bar. The slides then were washed for 30 minutes at 55°C in 0.2X SET, 0.1 % beta-mercaptoethanol. The tissues were dehydrated through an ethanol series (30% with 0.3M NH₄Ac, 50% with 0.3M NH₄Ac, 70% with 0.3M NH₄Ac, 80%, 90%, 95%, 100%) and allowed to air dry overnight.

Exposure and development: The following procedures were carried out in the dark or under a safety light: Dry slides were dipped in Kodak NTB-2 emulsion that had been diluted 1 : 1 with 600mM NH₄Ac at 42° C. Slides were allowed to dry for at least 1 hour. The slides were stored in black plastic light-tight boxes sealed with black electrical tape in a 4° C refrigerator free of radioactive sources.

The slides were developed in Kodak D-10 developer for 2.5 minutes at 15°C. Developing was stopped by immersion in 2% acetic acid for 30 seconds and the slides were then fixed in Kodak fixer for 6 minutes. Finally, slides were rinsed in tap water for half an hour to remove the fixer.

For counterstaining, the tissue sections were stained with Hematoxylin for 2 minutes, washed in distilled water for 2 minutes, and were dehydrated through another ethanol series as described above for tissue preparation. Finally, the slides were rinsed in xylene twice and covered permanently with a glass coverslip using permount (Fisher).

In all experiments, 4x10" cpm labeled riboprobe was used for each slide and signal was checked by developing a single slide at a time until desired exposures were obtained. Usually signal was detectable within 2-3 weeks.

Krox-20 transcript accumulation in normal murine embryos is observed first in rhombomere three followed by the later appearance of detectable transcripts in rhombomere five. At later stages, a decrease in accumulated signal occurs in rhombomere three, with the rhombomere five signal persisting beyond a point when rhombomere three signal is detectable. In rat embryos cultured in euglycemic rat serum the dynamic pattern observed for murine Krox-20 transcript accumulation is reproduced (Figure 1, a-f). In sections taken from embryos cultured for 16 hours (initiation site formation, 9- 10 somites), Krox-20 signal is detected in both rhombomeres three and five (Figure lb). Embryos cultured for 30 hours (14 somites) exhibit a decreased signal intensity in rhombomere three (Figure Id) and no signal is detected in rhombomere three in embryos at the 16 somite stage although a strong signal is still detected in rhombomere five (30 hour culture period, Figure If) .

Analysis of the Hoxb- 1 probe in embryos cultured in euglycemic media reveals a strong signal in rhombomere four spanning the time neural tube closure occurs (Figure 2 a-f) . Therefore, transient hindbrain segmentation and the accumulation of discrete domains of developmental control gene transcripts, milestones reflecting developmental and positional specification within the murine hindbrain, are faithfully replicated in rat embryos cultured in euglycemic rat serum.

A parallel examination of embryos cultured in hyperglycemic rat serum revealed the loss of visible segmentation within the hindbrain neuralepithelium defining distinct rhombomeres. In addition, prolapsed regions of the neural tube are observed and regions where the paired neural folds are abnormally adherent (Figure 1 , panels i and k). A number of other abnormalities are observed in these embryos; for example, otic pit formation is abnormal or absent in many exencepha lic embryos. Krox-20 transcripts exhibit the expected profile of transcript accumulation in rhombomeres three and five at the time of initiation site formation (9-10 somites, Figure lh) but persist abnormally in rhombomere three at the 14 and 16 somite stages (Figure 1, panel j and 1) .

Utilizing the Hoxb- 1 probe reveals that transcripts fail to accumulate to expected levels spanning the developmental interval from initiation site formation (Figure 2h) through the time neural tube closure is expected to be complete (Figure 2, panel j and 1). In the absence of the otic pit and rhombomeric segmentation, the rhombomere four region is identified by its proximity to the heart and the distance from the somites thus permitting the approximate regional assignment of the reduced signal to rhombomere four.

Thus, it appears that developmental control genes essential to the normal formation of the embryonic nervous system exhibit disruptions in their expression as a consequence of hyperglycemia. Krox-20 and hoxb- 1 show disruptions; and it is anticipated that other developmental control genes would be affected as well.

EXAMPLE 3; Folates Provide a Positive Control in the Bioassay

The majority of neural tube defects in humans are prevented by preconceptual folate dietary supplementation. Open cranial lesions (anencephaly) as well as open spinal lesions (spina bifida) are prevented by folate supplementation in both high-risk individuals and in the general population (see Laurence, et al., Brit. Med. J. 282: 1509-11 (1981); and A.E. Czeizel and I. Dudas, N. Engl. J. Med. 327: 1832-35 (1992)). However, because the folate doses that prevent ΝTDs are in significant excess of that needed to prevent dietary deficiency, the mechanism of action of folates in the prevention of birth defects remains in question. The bioassay of the present invention allows identification of agents which prevent ΝTDs. In this bioassay, rat embryos are explanted at 9.5 gestational days and cultured in euglycemic rat serum. Under these conditions, developmental progression continues comparable to that found in vivo.

Embryos were dissected from pregnant dams and distributed to each of three experimental classes: euglycemic culture, hyperglycemic culture, and hyperglycemic culture with a folinic acid pretreatment. The folinic acid pretreatment consisted of the addition of 20μ g/ml folinic acid one hour prior to the addition of excess glucose. At the end of the thirty hour culture period, embryos were scored for the presence of an open or closed neural tube. In euglycemic medium, 16.7% of the embryos exhibited exencephaly (5/30); in hyperglycemic medium 69.1 % exhibited exencephaly (29/42); in hyperglycemic medium with a folinic acid pretreatment, 29.4% of the embryos exhibited exencephaly (10/34). The difference in incidence of exencephaly between embryos cultured in hyperglycemic medium and those exposed to hyperglycemic medium containing folinic acid is statistically significant (X² with Yates correction factor, alpha=0.05, p=0.001 ).

Comparison of the incidence of exencephaly between embryos cultured in euglycemic medium versus embryos cultured in hyperglycemic medium plus folinic acid indicates that the incidence of exencephaly in these two categories is not different significantly (X² with Yates correction, alpha=0.01 , p=0.365). Thus, a one-hour folinic acid pretreatment prevented exencephaly in the cultured rat embryo.

The prevention of exencephaly achieved in the rat embryo study is comparable to that observed in human clinical trials using folate supplementation to prevent neural tube defects. Thus, a powerful in vitro model has been established in which the mechanism of action of folates in the prevention of neural tube defects can be investigated. In human studies, infants are scored at birth for the presence or absence of an open neural tube defect but these infants have not been assessed further to determine whether normal brain development occurred; therefore, the outcome of folate prevention efforts have not been explored fully in either clinical studies or in an animal model. To determine the outcome of folate pretreatment, a complete morphological assessment of each embryo was conducted (Table 1).

Table 1: Folinic acid pretreatment prevents exencephaly but produces primarily abnormal embryos.

ROSTRAL NEURALEPΓTHELIUM CLOSED:

Embryo D-glucose Normal Abnormal

Culture (mg/ml)/ Rostral Rostral exencephaly

(embryo Folinic acid Morph. Morphology number) (μg/ml)

Euglycemia 0 / 0 83.3% 0% 16.7%

(n= 30)

Hyperglycemia 4/0 9.5% 21.4% 69.1 %

(n= 42)

Hyperglycemia 4 / 20 17.6% 52.9% 29.4% and Folinic Acid (n= 34)

Embryo cultures were conducted as described in this Example and in Peng, Y., et al. , Int. J. Devi. Neuroscience 12:289-96 (1994), with modifications as described in this Example 3.

To test whether folinic acid (Sigma, Inc.) could suppress the open neural tube defect phenotype occurring in embryos cultured in hyperglycemic medium, standard culture conditions were employed as described above in Example 1 , except that folinic acid was added to the cultures medium at the outset of the culture period. Either 10 μg/ml or 20 μg/ml folinic acid from a stock solution (dissolved in sterile Hank's balanced salt solution; Sigma, Inc.) was added and embryos were cultured for one hour prior to the addition of D-glucose to construct hyperglycemic medium. Two embryos were co-cultured in 2 ml of media. To increase statistical rigor and limit variation due to genetic differences among litters, embryos from a single dam were distributed equally to each embryo culture group such that the products of a single pregnancy were not subjected to a single treatment protocol. Cultures were continued for a total of 30 hours and then a morphological assessment of embryonic development was conducted. Each embryo was examined for the degree of closure of the neuralepithelium noting whether the open NTD included the forebrain, midbrain, and hindbrain and whether caudal lesions were present. The neuralepithelium was examined further to determine whether it was prolapsed, abnormally adherent or twisted. The morphology of the optic and otic pits, pharyngeal arches, and tail were also noted as well as the number of somite pairs. In embryos with a closed neural tube, the size and morphology of the forebrain and midbrain tissue mass was evaluated and any extended hindbrain morphology was recorded. Selected embryos were photographed to document abnormal phenotypes using a Nikon SMZ-2T dissecting microscope.

To provide statistical validity and control for genetic variation between litters in the experiment presented above, embryos dissected from a single dam were distributed to all treatment categories thus providing littermates for each experimental and control class. To assess the significance of differences observed among experimental classes Chi^ statistical analysis was conducted using the Yates correction factor where appropriate.

Under euglycemic culture conditions, exencephaly was observed in 16.7% of the embryos, 83.3% exhibited a closed neural tube with normal morphology. Under hyperglycemic culture conditions 69.1 % of the embryos had exencephaly and 30.9% exhibited a closed neural tube. However, of those that did not exhibit exencephaly, more than two thirds showed significant morphological abnormalities and, in contrast to euglycemic controls, only one third of the embryos were normal (4/42). Thus, the hyperglycemic insult produces a spectrum of developmental damage with exencephaly as one outcome, but the hyperglycemic insult also produces a class exhibiting a closed neural lesion. Folate-pretreated embryos cultured in hyperglycemic serum showed a dramatic reduction in exencephaly with only 29.4% of the embryos affected. The class with the closed neural lesion was greatly increased, with 52.9% of the embryos exhibiting this phenotype. The numbers of embryos scored as normal approximately doubled from that in hyperglycemic media without folates: 17.6% of the embryos exhibited a closed neural tube with normal morphology. Pretreatment with folinic acid diminished developmental insult rather than preventing it— thus, the results with the folinic acid pretreatment protocol did not approximate the results found with culture in euglycemic serum; however, the decrease in exencephaly was substantial.

To assess further the extent of rescue due to folate pretreatment, the novel class of embryos with closed neural lesions were examined. A broad range of phenotypes was found (see representative embryos are depicted in Figure 3). The mildest abnormalities observed were the poor formation of the otic pit and the loss of rhombomeric segmentation in an otherwise normal embryo (not shown). The most extreme phenotypes observed included combinations of the following: abnormally adherent hindbrain neuralepithelium with prolapsed regions that failed to fuse (Figure 3e), malformation or loss of the otic pit (Figure 3e and 3f), twisting of the neuralepithelium (Figure 3f) , reduction of the forebrain and midbrain tissue mass (Figures 4b and 4c), reduction of the pharyngeal arches to a single enlarged arch (Figure 3b and 3c), an extended hindbrain morphology (Figure 4c), and the failure to rotate the tail (Figures 4b a n d 4c) .

An identical spectrum of defects is found in embryos treated with hyperglycemia alone but their relative numbers are much lower, as the majority of these embryos exhibit exencephaly. The complete absence of these defects in embryos cultured in euglycemic media suggest that these abnormalities derive from the hyperglycemic insult. Thus, the bioassay has positively identified folates as compounds which can blunt the impact of hyperglycemia, preventing diabetic complications.

EXAMPLE 4: Transcript Accumulation Profiles of Krox-20 in Folate-Pretreated Embryos

The striking similarity between the overall morphology of exencephalic embryos and those bearing the closed neural lesion (reduction in forebrain and midbrain tissue mass, a single enlarged pharyngeal arch, etc.) suggests that these lesions differ primarily in severity and that both comprise neural tube defects arising from the same process: a teratogenic exposure to glucose . It appears that disruptions in the same patterning and adhesive events that occur when neural tube closure fails have also occurred in these morphologically abnormal embryos. To test this hypothesis and to determine whether patterning defects are evident in embryos with the abnormal morphology phenotype, tissue sections were prepared from folate-pretreated embryos and hybridized in situ with both the Hoxb- 1 and Krox-20 murine probes, using the protocol described in Example 2. Krox-20 encodes a developmentally-regulated transcription factor bearing a zinc-finger motif that has been implicated recently in the regulation of homeobox gene expression (Nonchev, S ., et al., Deve lopment 122 :543-54 ( 1996)) . Hoxb- 1 encodes a homeodomain transcription factor that is also essential to the normal execution of the development of the cranial nervous system (Goddard, et al., Development ( 1996)) .

An embryo with normal morphology (Figure 4a) exhibits residual Krox-20 signal in rhombomere three not ordinarily observed at the sixteen somite stage (Figure 4d). An embryo with moderately abnormal morphology shows a more abnormal Krox-20 signal, with rhombomere three and five signals approximately equivalent (Figure 4b). A grossly abnormal embryo exhibits a Krox-20 molecular signature indistinguishable from that found in exencephalic embryos (Figure 4f, see also Peng, Y. and K. Fechtel, Teratology (in press)). In the examination of over ten different embryos cultured in hyperglycemic medium with a folinic acid pretreatment and subsequently hybridized i n situ with the Krox-20 probe, no embryo exhibited the expected molecular phenotype for the observed developmental stage (somite number). A consistent trend is observed that as morphological disruption increases, the underlying molecular evidence that patterning events are disrupted is also increased; a result that is consistent and reproducible. Therefore, as seen in Example 3, the folinic acid pretreatment protocol employed appears to have diminished the impact of the teratogenic insult at the molecular level in the majority of embryos, but not eliminated it.

Examination of the Hoxb- 1 transcript accumulation profile in three folate pretreated embryos reveals the normalization of transcript accumulation profiles as well. In each case, significant Hoxb- 1 transcripts were detected where in the same culture conditions without folate pretreatment little or no transcript accumulation is observed. This finding is illustrated in Figure 5. Although the representative embryo (Figure 5a) retains morphological abnormalities (irregular neuralepithelium), a significant Hoxb-1 signal is detected (Figure 5b).

These analyses of developmental control gene expression profiles clearly establish that the folate pretreatment blunted the impact of hyperglycemia at the molecular level as well and demonstrate the utility of molecular markers indicating the achievement of developmental landmarks in the bioassay demonstrating the efficacy of agents that blunt the biochemical process leading to diabetic complications.

EXAMPLE 5: Identification of Pterinaldehydes as Compounds Effective in Preventing

Neural Tube Defects Resulting from Hyperglycemia

Employing the bioassay of the present invention, a class of compounds known as pterins has been identified as being effective in preventing diabetic complications resulting from hyperglycemia. Until the present invention, pterins had not been linked previously to diabetes, either theoretically or as inhibitors of currently-studied processes. Pterins have very low toxicity and therefore are excellent preventative therapeutics, exhibiting efficacy in the prevention of diabetic complications, atherosclerosis, birth defects, and the recurrence of certain cancers . The model pterin employed in this example is pterinaldehyde, a photolytic degradation product of folic acid that can be synthesized readily from folic acid. Pterinaldehyde prevented NTDs in the rat embryos assay showing markedly better results than folates with at least a two-fold better outcome with an approximately tenfold lower dose.

Experimental Design: Embryo cultures were executed as in Example 3; however, instead of folinic acid added one hour prior to establishing hyperglycemic cultures, 2, 4, or 6 μ g of pterinaldehyde was added to the rat embryo cultures to test for the ability of this model pterin to prevent NTDs in the bioassay. Pterinaldehyde was synthesized from folic acid using the protocol of Thijsen (( 1973)). All procedures were carried out in a fume hood. For each synthesis, 12g Bromine (Sigma, Inc.) was added into 30ml of a 40% hydrogen bromide solution (Sigma Chemical Co., Inc.) in a flask and the solution was thoroughly mixed using a magnetic stir bar. Folic acid (6.6 g) was added gradually into the mixture with continuous stirring. When the folic acid was dissolved, the solution was heated to 100°C using a heating mantle for 2 hours with continuous stirring with a magnetic bar. The resulting precipitate, the primary pterinaldehyde, was collected by filtration and washed thoroughly with water. The precipitate was extracted with hot acetone (Sigma, Inc.). This was carried out by mixing the precipitate with acetone in a beaker with continued heating on a hot plate preheated to approximately 60° C. The precipitate was collected by filtration 30 minutes later. The wash procedure was repeated twice. The yellow precipitate was reported to be pterinaldehyde, and it was dried in a vacuum oven at 50° C overnight. Pterinaldehyde was kept in a brown bottle, wrapped with aluminum foil, and stored at 4° C. In a typical synthesis reaction, the yield was usually one gram from 6.6 grams of folic acid. For use in whole embryos culture experiments, the crude, synthetic pterinaldehyde was dissolved in Hank's solution at 125(g/ml as a stock solution, and stored at 4°C in a glass bottle wrapped with aluminum foil prior to use.

High performance liquid chromatography (HPLC) was used to verify that commercial preparations (Fluka and Sigma brand) of folinic acid exhibit varying levels of contamination of a substance that co-elutes with the synthesized pterinaldehyde.

The HPLC analysis was performed according to the method of

Ferone and Spector ( 1984)). The HPLC used was a Waters 600S controller 626 pump with in-line UV-Vis detection using a Waters

996 photodiode array. An isocratic buffer system was employed. The C _j g column (Vydac, Inc.) was equilibrated with the elution buffer (5 mM tetrabutylammonium phosphate and 30% methanol in 1 mM ammonium phosphate, pH 6.5) and samples w ere injected in a total volume of 50 μl with the concentrations ranging from 20 to 30 μg/ml. Each run was executed for 40 minutes and the absorbance was monitored at 260nm. In this experiment, Sigma folinic acid showed a greater level of contamination with a substance that co-elutes with pterinaldehyde. Pterinaldehyde is reported to elute at approximately 5 minutes in this HPLC method these results match the reported findings. However, although th e reported synthesis indicates the production of a high purity reagent, HPLC analysis revealed a number of contaminant peaks arising during the synthesis protocol.

Table 2 demonstrates that a substantially lower dose of the pterinaldehyde synthesis product (2-6μg of which 30-50% is estimated to be pterinaldehyde) produces a substantially better outcome than is found with 20 μg folinic acid with a reduction of open neural tube defects equal to that found in euglycemic culture conditions achieved by the 4 μg dose. Equally important, th e numbers of normal embryos produced exceeded that obtained for folinic acid with 54.5% or 55% normal embryos wh en pterinaldehyde is utilized as the preventative agent. At th e highest dose employed, 6 μg, solubility problems may have b een responsible for the lower than expected prevention of the open neural tube defects. Table 2: Pterinaldehyde preparation prevents ex e n c ep h a ly in an in vitro hyperglycemia model

Culture D-glucose Nι armal Abn ormal

Conditions ( m g/m 1 ) / M orph . Morph. exencephaly

(embry o Pterina ildehyde ( % ) ( % ) ( % ) number) (μg/ml^')

Euglycemia 0/0 72.7 4.5 2.7

(n= 22)

Hyperglycemia 4/0 16.7 16.7 66.6

(n= 24)

Hyperglycemia 4 / 2 21.7 30.4 47.8

Pterinaldehyde

(n= 23)

Hyperglycemia 4/4 54.5 * =* 22.7 22.7 *

Pterin aldehyde

(n= 22)

Hyperglycemia 4/6 55.0** 5 40 * *

Pterinaldehyde

(n= 20

**significantly different from hyperglycemia by Chi analysis (alpha = 0.05, p < 0.01)

This example demonstrates the utility of the bioassay in which an unknown compound was demonstrated to show efficacy as a preventative agent in complications arising as a consequence of hyperglycemia and indicates that pterins as well as folates prevent or ameliorate birth defects arising in a model of maternal diabetes.

Chemical analysis of the pterinaldehyde synthetic product used in the prevention assay indicates that it contains a mixture of other folate derivatives with pterinaldehyde the primary component by HPLC analysis. Thus, although it is not known which HPLC peak is responsible for the majority of the bioactivity, it is clear that folate derivatives show greater efficacy than folinic acid using the explanted rat embryo assay in which hyperglycemia-induced neural tube defects are prevented. The dramatic prevention effect demonstrated by folinic acid and the pterinaldehyde synthetic product (presumably pterinaldehyde) warrants the use of pterinaldehydes and folate derivatives as therapeutic agents in the prevention of diabetic complications such as neural tube defects and atherosclerosis. Substantial clinical evidence links folates with the prevention of neural tube defects and with a reduced risk of cardiovascular disease. The identification of agents with greater efficacy than folates fulfills a need in the art. Further, a relationship between folates, atherosclerosis and neural tube defects has not been explored previously in diabetic human subjects and the present invention is the first disclosure linking folates with the prevention of diabetic complications.

Any patents or publications mentioned in this specification are indicative of the levels of those skilled in the art to which the invention pertains. Further, these patents and publications are incorporated by reference herein to the same extent as if each individual publication was specifically and individually indicated to be incorporated by reference.

One skilled in the art will appreciate readily that the present invention is well adapted to carry out the objects and obtain the ends and advantages mentioned, as well as those objects, ends and advantages inherent herein. The present examples, along with the methods , procedures, treatments, molecules, and specific compounds described herein are presently representative of preferred embodiments, are exemplary, and are not intended as limitations on the scope of the invention. Changes therein and other uses will occur to those skilled in the art which are encompassed within the spirit of the invention as defined by the scope of the claims.

Claims

WHAT IS CLAIMED IS:

1 . An in vitro bioassay to determine agents which can blunt the deleterious effects of hyperglycemia in an animal or its offspring, comprising the steps of: selecting an agent to be tested; dissecting rat embryos from pregnant dams; culturing said dissected rat embryos in hyperglycemic rat serum and said agent to be tested for a pre-determined time; and examining said cultured rat embryos to determine if abnormalities exist, wherein if abnormalities exist, said agent does not blunt deleterious effects of hyperglycemia, and if abnormalities do not exist, said agent does blunt deleterious effects of hyperglycemia.

2. The in vitro bioassay of claim 1 , wherein said dissecting step is performed at about gestational day 7-12 of said rat embryos.

3. The in vitro bioassay of claim 1 , wherein said dissecting step is performed at about gestational day 9-10 of said rat embryos.

4. The in vitro bioassay of claim 1 , wherein said predetermining time for said culturing step is about 20-40 hours.

5. The in vitro bioassay of claim 4, wherein said predetermining time for said culturing step is about 30 hours.

6. The in vitro bioassay of claim 1 , wherein said examining step includes observing an appearance of a neural tube; pharyngeal arches; otic and optic pits; tail flexure; positioning of hindbrain, midbrain, forebrain or heart; state of neural epithelium; or somite number of said embryos.

7. The in vitro bioassay of claim 1 , wherein said examining step includes evaluating transcript accumulation profiles of developmental control genes.

8. The in vitro bioassay of claim 7, wherein said developmental control genes are Krox-20 and Hoxb-1.

9. The in vitro bioassay of claim 1 , wherein folates, folate derivatives or pterins are added to said hyperglycemic rat serum as a positive control.

10. An in vivo bioassay to determine agents which can blunt the deleterious effects of hyperglycemia in an animal or its offspring, comprising the steps of: selecting an agent to be tested; inducing insulin-dependent diabetes in dams following a glucose challenge; treating said dams with insulin; establishing pregnancy in said dams; withdrawing insulin from said dams; administering glucose and said agent to be tested in a pharmacologically acceptable carrier to said dams; dissecting embryos from said dams; and examining said cultured rat embryos to determine if abnormalities exist, wherein if abnormalities exist, said agent does not blunt deleterious effects of hyperglycemia, and if abnormalities do not exist, said agent does blunt deleterious effects of hyperglycemia.

1 1 . The in vivo bioassay of claim 10, wherein said dissecting step is performed at about gestational day 7-12 of said rat embryos.

12. The in vivo bioassay of claim 10, wherein said dissecting step is performed at about gestational day 9-10 of said rat embryos.

13. The in vivo bioassay of claim 10, wherein said treating step is performed by injecting said dams with said insulin three times per day.

14. The in vivo bioassay of claim 10, wherein said withdrawing step is performed 1 -2 days before said dissecting step .

15. The in vivo bioassay of claim 10, wherein said examining step includes observing an appearance of a neural tube; pharyngeal arches; otic and optic pits; tail flexure; positioning of hindbrain, midbrain, forebrain or heart; state of neural epithelium; or somite number of said embryos.

16. The in vivo bioassay of claim 10, wherein said examining step includes evaluating transcript accumulation profiles of developmental control genes.

17. The in vivo bioassay of claim 16, wherein said developmental control genes are Krox-20 and Hoxb-1.

1 8. The in vivo bioassay of claim 10, wherein folates, folate derivatives or pterins are administered to said dams as a positive control.

19. A compound effective in preventing complications resulting from hyperglycemia, selected from the group of folates, folate derivatives and pterins.

20. A compound effective in preventing cardiovascular disease and neural tube defects, selected from the group of folates, folate derivatives and pterins.