Prediction and prophylactic treatment of type 1 diabetes

Abstract

An in vitro method for predicting the onset of type 1 diabetes (T1D) in a subject, comprises the steps of: (a) measuring the concentration of at least one amino acid, amino acid derivative or amino acid metabolite in a biological sample taken from the subject; (b) determining the subject's HLA genotype; (c) assigning the subject's genetic risk of developing T1D on the basis of the subject's HLA genotype; (d) combining the information obtained in step (a) with the information in step (c); and (e) predicting the likelihood of onset of T1D based upon the combination of step (d). The diagnostic method can be used to select target subjects for T1D prophylactic treatment, and as part of a T1D preventative treatment regime for neonates having a likelihood of developing childhood T1D.

Claims

1. An in vitro method for delaying the onset of type 1 diabetes (T1D) in a human identified as indicative for the future onset of T1D, said method comprising administering to said human a composition comprising free carnitine and/or acylcarnitine (L-carnitine), in an amount sufficient to restore free carnitine, acylcarnitine levels to within the normal range of 50 mol/L or more for a high risk human leukocyte antigen (HLA) genotype, or to within the normal range of 40 mol/L or more for a moderate or low risk HLA genotype, and wherein said human is identified as indicative for the future onset of T1D by: (a) measuring the total concentration of amino acids alanine, arginine, aspartate, citrulline, glycine, glutamate, glutamine, leucine, isoleucine, methionine, ornithine, phenylanaline, proline, tyrosine and valine via tandem mass spectrometry in a biological sample taken from said human within 3 days of the human's birth, wherein the biological sample is a blood sample; (b) measuring the concentration of total carnitine via tandem mass spectrometry in a sample taken from said human within 3 days of the human's birth; (c) determining the human's HLA genotype via PCR; (d) assigning the human's genetic risk of developing T1D on the basis of the human's HLA genotype; (e) combining the information obtained in (a) with the information in (d); (f) combining the information obtained in (b) with the information in (d); and (g) predicting the likelihood of onset of T1D based upon the combination of (e) and (f); wherein in (d) the genetic risk is assigned as high, moderate or low according to the human's HLA genotype; wherein the high risk HLA genotype comprises the genetic markers DRB1*03/*04 (not 0403) and DQB1 0302; wherein a moderate/medium risk HLA genotype comprises the genetic markers DRB1*04 (not 0403)/*04 (not 0403), DQB1 0302, DRB1*04 (not 0403)/X, DQB1 0302/DQB1 (not 0602-3), DRB1*03/*03, DRB1*03/X, and DQB1 (not 0602-3, 0301, or 0503); wherein X is not DRB1*03, *04, or 0403; and wherein the low risk HLA genotype comprises any combination of genetic markers not specified as high or moderate risk; wherein a total amino acid concentration of less than approximately 1200 mol/L and a total carnitine concentration of less than approximately 50 mol/L in combination with a high risk HLA genotype as identified in (g) is indicative of the future onset of T1D; and wherein in (g) a total amino acid concentration of less than approximately 1000 mol/L and a total carnitine concentration of less than approximately 40 mol/L independent of the HLA genotype as identified in (g) is indicative of the future onset of T1D.

2. The method of claim 1, wherein in (g) a total amino concentration of approximately 500-1200 mol/L, and a total carnitine concentration of approximately 20-50 mol/L in combination with a high risk HLA genotype as identified in (g) is indicative of the future onset of T1D.

3. The method of claim 1, wherein in (g) a total amino acid concentration of approximately 300-1000 mol/L, and a total carnitine concentration of less than approximately 40 mol/L in combination with a moderate or low risk HLA genotype as identified in (g) is indicative of the future onset of T1D.

4. The method of claim 1, wherein in (g) a total amino acid concentration of approximately 300-1000 mol/L, and a total carnitine concentration of approximately 10-40 mol/L in combination with a moderate or low risk HLA genotype as identified in (g) is indicative of the future onset of T1D.

5. The method of claim 1, wherein (c) is carried out by genetically screening a portion of said sample used in (a).

6. The method of claim 1, wherein (b) is carried out on the sample used in (a).

Description

(1) The invention is further illustrated by the accompanying drawings in which:

(2) FIG. 1 is a graph showing the relationship between the total concentration of amino acids in blood samples taken from T1D patients and non-diabetic negative controls, in comparison to their HLA genotype risk category. The Medium/moderate and low HLA risk categories are grouped together. The boxes indicate the 75.sup.th percentile (top of box) and the 25.sup.th percentile (bottom of box).

(3) FIG. 2 is a graph showing the relationship between the total concentration of carnitine and acylcarnitine in blood samples taken from T1D patients and non-diabetic negative controls, in comparison to their HLA genotype risk category. Medium/moderate and low HLA risk categories are grouped together. The boxes indicate the 75.sup.th percentile (top of box) and the 25.sup.th percentile (bottom of box).

(4) FIG. 3 is a graph showing the relationship between the total concentration of amino acids in blood samples taken from T1D patients and non-diabetic negative controls, in comparison to their HLA genotype risk category, plotted in the 95% confidence interval. Medium/moderate and low HLA risk categories are grouped together.

(5) FIG. 4 is a graph showing the relationship between the total concentration of carnitine and acylcarnitine in blood samples taken from T1D patients and non-diabetic negative controls, in comparison to their HLA genotype risk category, plotted in the 95% confidence interval. Medium/moderate and low HLA risk categories are grouped together.

EXAMPLE

(6) (i) Methods

(7) A prospective study was conducted to determine the relevance of concentration of circulating amino acids, amino acid derivatives, amino acid metabolites, free carnitine and acylcarnitine to the future development of type 1 diabetes (T1D).

(8) Diabetic patients were identified from within the Unit of Diabetology at the Bambino Ges Paediatric Hospital in Rome. Controls were selected from 1650 children, who were HLA typed at birth for genetic susceptibility to develop T1D (Buzzetti R. et al. 2004, Diabetes Metab Res Rev. 20(2): 137-43).

(9) As a criteria of inclusion, children had be born in Lazio between January 2000 and December 2002 and T1D had to be developed by the age of 4 years. These criteria were adopted in order to: 1) have the possibility of retrieving dry blood spot samples; 2) reduce the time elapsing between birth and conducting the assays; and 3) identify controls matched for age, sex and genetic HLA risk categories (i.e. high, moderate/medium and low), defined as previously described in Buzzetti R. et al. 2004, supra.

(10) Overall, 11 diabetic children fulfilled the selection criteria and 44 matched controls were identified. These 11 diabetic children represent nearly all of the children who have developed T1D in the first 4 years of life from a sample size of 150,000 children born in the Lazio region over the relevant time period. Information regarding the subjects used in the study is provided in Table 1.

(11) Informed consent was obtained from parents of all children.

(12) Information regarding birth weight, gestational age at birth and feeding in the first week of life was recorded.

(13) Blood dry spots were retrieved from the two centers in charge of neonatal screenings (e.g. congenital hypothyroidism and phenylketonuria) in the same region. All spots were collected within the first days after birth (i.e. during the neonatal period) and stored at room temperature until the analysis. Once samples had been identified, they were coded and no information regarding T1D status was available before the results of the study had been obtained.

(14) For each blood spot, concentrations of 15 amino acids (in 13 categories), along with the concentrations of total carnitine (TC), free carnitine (FC) and acylcarnitine (AC) were measured and recorded (see Table 2). Internal AC and amino acid standards were provided by Cambridge Isotopes Laboratories, Andover, Mass., USA.

(15) Carnitine and amino acids were analyzed on filter paper (W903) blood spots as butyl esters. The butylated samples were tested with a tandem mass spectrometer (Sciex API 365, PE Sciex Instrument, Concord, ON, Canada). This mass spectrophotometer is routinely used for neonatal mass screening. Quality control of the methods for measuring amino acids, carnitine and acylcarnitine was performed by Dr Piero Rinaldo, (Mayo Clinic, Ronchester, Minn., USA).

(16) TC, FC, AC and AC/FC ratio were analyzed. Amino acids were also studied as the algebraic sum of essential, non-essential and total amino acid concentrations. The Mann-Whitney U-Test was used for comparison of continuous variables. Frequencies of qualitative variables were evaluated using Fisher's exact test. A two-tailed P value of less than 0.05 was considered to be significant. All analyses were performed using SPSS 11.5.1 for Windows Package (LEAD Technologies, Inc).

(17) (ii) Results

(18) Referring to Table 1, among the 11 diabetic children, 4 were found to possess the high risk HLA genotype for developing T1D; 6 were found to have the medium/moderate risk HLA genotypes; and 1 had a low risk HLA genotype.

(19) The median age at the onset of T1D was 2.7 years, and onset age ranged from 1.1 to 3.8 years.

(20) No differences were seen in reference to gestational age at birth, birth weight and type of feeding in the first week of life (P=0.438, P=0.408 and P=0.522, respectively). Thus, in this sample it did not appear that feeding on formula milk or breast milk was relevant.

(21) The time elapsing from the collection of blood spots to the performance of the assays was comparable in the two groups (P=0.191), so it is considered that any slight differences were insignificant.

(22) Turning to Table 2 and FIGS. 1 and 3, it can be seen that the diabetic patients displayed lower levels of all of the measured amino acids than did the non-diabetic controls. Moreover, the concentrations of the amino acids Ala, Gly, Glu/Gln, Leu/Ile, Orn, Phe and Pro were seen to be significantly lower in diabetic subjects than in the non-diabetic control subjects. Overall, total amino acid concentrations were significantly lower in diabetic patients than in controls, as were the levels of essential amino acids and non-essential amino acids. Although the concentration of the essential amino acids shown in Table 2 does not include that of arginine (which is essential in neonates), if arginine is included in the essential amino acid subgroup, rather than in the non-essential subgroup, the above observation regarding the relative levels of amino acids remains true.

(23) FIG. 1 shows the total amino acid concentrations measured in blood samples of either T1D patients or non-diabetic controls, depending on their HLA genotype risk category. The boxes indicate the 25th to the 75th percentile. In addition to the lower relative concentration of amino acids found in the samples of T1D patients, the range of total amino acid concentrations measured appears to be narrower in the T1D patients.

(24) FIG. 3 shows the 95% confidence interval (95% CI) of the total amino acid concentrations measured in blood samples of either T1D patients or non-diabetic controls, displayed according to the subject's HLA genotype risk category. Again, it can be seen that the amino acid concentrations are significantly lower in the T1D patients than in the non-diabetic controls.

(25) Also from Table 2 and FIGS. 2 and 4, it can be seen that circulating TC, FC and AC concentrations were significantly lower in diabetic patients compared to the non-diabetic controls. However, essentially no differences were detected in the AC/FC ratio.

(26) FIG. 2 shows the total carnitine concentrations measured in blood samples of either T1D patients or non-diabetic controls, depending on their HLA genotype risk category. The boxes indicate the 25th to the 75th percentile. Again, in addition to the lower relative amount of carnitine found in the samples of T1D patients, the range of total carnitine concentrations measured appears to be narrower in the T1D patients.

(27) FIG. 4 shows the 95% confidence interval (95% CI) of the total carnitine concentrations measured in blood samples of either T1D patients or non-diabetic controls, displayed according to the subject's HLA genotype risk category. Again, it can be seen that the carnitine concentrations are significantly lower in the T1D patients than in the non-diabetic controls.

(28) The data used in FIGS. 1 and 2 is displayed in Table 3 and the data used in FIGS. 3 and 4 is displayed in Table 4.

(29) (iii) Conclusions

(30) This is the first demonstration that neonates who go on to develop T1D early in life (i.e. in the first 4 years) display reduced concentrations of amino acids and of carnitine during the neonatal period.

(31) Since this investigation was carried out retrospectively on dried blood spots stored at room temperature, as a control, 80 dried blood spots taken from children of the same region, and which have undergone similar storage conditions in the same centers for two and four years, respectively, have been tested using the same tandem mass spectrometer used for the Example.

(32) Values of amino acids and of TC, FC and AC, were measured and found to be very similar in the two sample sets (inventors' unpublished results). In the Example described, since the time elapsing between blood collection and the conduction of the assays was similar between diabetics and controls, and was carried out in the time range of 2-4 years, the differences in amino acid and carnitine concentrations noted between diabetics and non-diabetic controls can be considered not to be due to the latency between blood collection and sample testing.

(33) For the reasons already discussed with regard to Table 1, it is also possible to exclude the differences in the type of feeding, because the proportion of children whom were breast-fed or whom were fed formula milk was comparable between the two groups.

(34) It is therefore concluded that children who later developed T1D early in life have reduced levels of circulating amino acids and carnitine at birth or soon after birth (i.e. during the neonatal period). Thus, low levels of amino acids and of carnitine may be viewed as additional markers for predicting the future onset of T1D in children, irrespective of their perceived genetic susceptibility.

(35) The determination of these compounds is easy to do and can be routinely applied in population screenings. Hence, new perspectives for the testing of all newborn children, to predict the likely onset of T1D, and if necessary commence a suitable prophylactic treatment regime are apparent.

(36) TABLE-US-00001 TABLE 1 Characteristics of diabetic children and matched controls Diabetic Control Patients children Significance (N = 11) (N = 44) P M:F 7:4 28:16 HLA risk category high 4 16 moderate 6 24 low 1 4 Gestational age at birth 39 (37-41) 39 (30-41) 0.438 (weeks) Birth weight (Kg) 3.5 (3.0-4.1) 3.4 (1.6-4.8) 0.408 Feeding during the 1.sup.st week only breast milk 6 26 0.522 formula or mixed 5 18 Age at onset of T1D (yrs) 2.7 (1.1-3.8) Age at metabolic 3.9 (2.5-5.0) 3.2 (1.6-4.9) 0.191 evaluation (yrs)
Values are expressed as number of cases or as median (range).

(37) TABLE-US-00002 TABLE 2 Carnitine and amino acid concentrations (mol/L) in blood dry spots collected within the first three days of life Diabetic Patients Control children Signifi- (N11) (N44) cance P Total carnitine 29.9 (14.8-38.9) 39.5 (19.7-96.1) 0.004 Free carnitine 19.0 (6.8-25.6) 24.7 (11.7-68.9) 0.009 Acylcarnitine 10.7 (7.8-16.3) 16.5 (8.0-27.2) 0.009 Acyl/free ratio 0.7 (0.4-1.2) 0.6 (0.3-1.2) 0.556 Alanine (Ala) 82.5 (50.7-174.4) 140.8 (50.1-526.7) 0.037 Arginine (Arg) 15.1 (10-34.6) 17.4 (7.3-49.0) 0.599 Aspartate (Asp) 35.9 (15.8-70.4) 42.9 (18.0-152.9) 0.461 Citrulline (Cit) 4.6 (2.9-13.3) 6.8 (2.4-22.1) 0.064 Glycine (Gly) 80.9 (49.8-131.2) 122.1 (66.0-354-7) 0.002 Glutammate/ 223.8 (114.4-332.1) 314.0 (211.2-578.1) 0.002 Gluttammine (Glu/Gln) *Leucine/ 60.9 (42.2-84.6) 86.9 (51.5-191.2) <0.001 lsoleucine (Leu/Xle) *Methionine 3.6 (1.9-8.2) 4.4 (1.3-10.9) 0.326 (Met) Ornitine (Orn) 6.7 (4.0-16.8) 11.0 (3.4-34.8) 0.001 *Phenylalanine 15.2 (3.9-29.8) 24.9 (12.4-54.8) 0.001 (Phe) Proline (Pro) 45.3 (24.4-73.7) 73.9 (34.2-128.5) 0.002 Tyrosine (Tyr) 26.4 (14.1-65.2) 32.3 (13.7-80.3) 0.323 *Valine (Val) 59.8 (29.8-150) 67.8 (10-250) 0.192 Essential 134 (91-242) 177 (120-426) 0.003 aminoacids Non essential 549 (424-852) 781 (465-1805) 0.003 aminoacids Total 677 (515-1016) 954 (618-2230) 0.003 aminoacids 5 Values are expressed as median (range) *Essential amino acids (excluding arginine)

(38) TABLE-US-00003 TABLE 3 Total amino acid and carnitine concentrations measured in samples taken from subjects with either the high risk HLA genotype or with the moderate or low risk HLA genotypes (as plotted in FIGS. 1 and 2 respectively) moderate or moderate or high risk high risk low risk low risk diabetics controls diabetics controls (Nr = 4) (Nr = 16) (Nr = 7) (Nr = 28) Total amino acids (mol/L) maximum 1016.4 2231 909.6 1628.3 75.sup.th percentile 996.4 1543 677.2 1191.7 median 836.8 1229.6 673.3 849 25.sup.th percentile 735.8 960.5 520.5 715.9 minimum 735.2 618.5 515.6 632.2 Total carnitine (mol/L) maximum 38.9 69.6 31.8 96.1 75.sup.th percentile 38.4 57.1 31.78 44.8 median 35.4 49.2 26.4 37.8 25.sup.th percentile 30.7 31.8 18.9 30.2 minimum 29.7 19.7 14.8 25.3

(39) TABLE-US-00004 TABLE 4 Total amino acid and carnitine concentrations measured in samples taken from subjects with either the high risk HLA genotype or with the moderate or low risk HLA genotypes (as plotted in FIGS. 3 and 4 respectively) moderate or moderate or high risk high risk low risk low risk diabetics controls diabetics controls (Nr = 4) (Nr = 16) (Nr = 7) (Nr = 28) Total amino acids (mol/L) 95% CI upper limit 1082.9 1493.4 774.5 1039.7 95% CI lower limit 629.8 1035.6 523.2 833.2 95% CI midpoint 856.3 1264.5 648.8 936.5 Total carnitine (mol/L) 95% CI upper limit 41.3 53.2 31.3 46.9 95% CI lower limit 28.4 37.1 19.1 35.1 95% CI midpoint 34.9 45.1 25.2 41

REFERENCES CITED

(40) Akerblom H K, Vaarala O, Hyoty H, Ilonen J, Knip M., Environmental factors in the etiology of type 1 diabetes. (2002), Am J. Med. Genet., May 30; 115(1): 18-29. Virtanen S M, Knip M., Nutritional risk predictors of beta cell autoimmunity and type 1 diabetes at a young age. (2003) Am. J. Clin. Nutr., Dec; 78(6): 1053-67. Ziegler A G, Schmid S, Huber D, Hummel M, Bonifacio E., Early infant feeding and risk of developing type 1 diabetes-associated autoantibodies. (2003) J. Am. Med. Assoc., October 1; 290(13): 1721-8. Redondo M J, Fain P R, Eisenbarth G S. Genetics of type 1A diabetes.(2001) Recent Prog. Horm. Res. 56: 69-89. Ilonen J, Sjoroos M, Knip M, Veijola R, Simell O, Akerblom H K, Paschou P, Bozas E, Havarani B, Malamitsi-Puchner A, Thymelli J, Vazeou A, Bartsocas C S. (2002) Estimation of genetic risk for type 1 diabetes. Am. J. Med. Genet. 115(1): 30-6. Mimbacas A, Perez-Bravo F, Hidalgo P C, Javiel G, Pisciottano C, Grignola R, Jorge A M, Gallino J P, Gasagoite J, Cardoso H. (2003) Association between diabetes type 1 and DQB1 alleles in a case-control study conducted in Montevideo, Uruguay. Genet. Mol. Res. 2(1): 29-35. Lambert A P, Gillespie K M, Thomson G, Cordell H J, Todd J A, Gale E A, Bingley P J. Absolute risk of childhood-onset type 1 diabetes defined by human leukocyte antigen class II genotype: a population-based study in the United Kingdom. (2004) J. Clin. Endocrinol. Metab. 89(8): 4037-43. Buzzetti R, Galgani A, Petrone A, Del Buono M L, Erlich H A, Bugawan T L, Lorini R, Meschi F, Multari G, Pozzilli P, Locatelli M, Bottazzo G, Di Mario U., Genetic prediction of type 1 diabetes in a population with low frequency of HLA risk genotypes and low incidence of the disease (the DIABFIN study). (2004), Diabetes Metab. Res. Rev., Mar-Apr; 20(2): 137-43. Galgani A, Petrone A, Spoletini M, Hodge A, Del Buono M L, Locatelli M, Buzzetti R and the Diabfin Study Group. HLA class II typing in newborns reveals a low frequency of the DRB1*04 allele and a high frequency of DRB1*11 allele in three regions of continental Italy. (2004) Hum. Immunol. 65(4): 366-72.