Regulations and Funding to Create Enterprise Architecture for a Nationwide Health Data Ecosystem

The COVID-19 pandemic highlighted the United States' lack of a nationwide infrastructure for collecting, sharing, and using health data, especially for secondary uses (e.g., population health management and public health). The federal government is taking several important steps to upgrade the nation's health data ecosystem-notably, the Centers for Disease Control and Prevention's Data Modernization Initiative and the Office of the National Coordinator for Health Information Technology's Trusted Exchange Framework and Common Agreement. However, substantial barriers remain. Inconsistent regulations, infrastructure, and governance across federal and state levels and between states significantly impede the exchange and analysis of health data. Siloed systems and insufficient funding block effective integration of clinical, public health, and social determinants data within and between states. In this analytic essay, we propose strategies to develop a nationwide health data ecosystem. We focus on providing federal guidance and incentives to develop state-designated entities responsible for the collection, integration, and analysis of clinical, public health, social determinants of health, claims, administrative, and other relevant data. These recommendations include a regulatory clearinghouse, federal guidance, model legislation and templated regulation, funding to incentive enterprise architecture, regulatory sandboxes, and a 3-pronged research agenda. (Am J Public Health. 2024;114(2):209-217. https://doi.org/10.2105/AJPH.2023.307477).

Human blood and marrow side population stem cell and Stro-1 positive bone marrow stromal cell numbers decline with age, with an increase in quality of surviving stem cells: correlation with cytokines

Hematological deficiencies increase with aging leading to anemias, reduced hematopoietic stress responses and myelodysplasias. This study tested the hypothesis that side population hematopoietic stem cells (SP-HSC) would decrease with aging, correlating with IGF-1 and IL-6 levels and increases in bone marrow fat. Marrow was obtained from the femoral head and trochanteric region of the femur at surgery for total hip replacement (N=100). Whole trabecular marrow samples were ground in a sterile mortar and pestle and cellularity and fat content determined. Marrow and blood mononuclear cells were stained with Hoechst dye and the SP-HSC profiles acquired. Marrow stromal cells (MSC) were enumerated flow cytometrically employing the Stro-1 antibody, and clonally in the colony forming unit fibroblast (CFU-F) assay. Plasma levels of IGF-1 (ng/ml) and IL-6 (pg/ml) were measured by ELISA. SP-HSC in blood and bone marrow decreased with age but the quality of the surviving stem cells increased. MSC decreased non-significantly. IGF-1 levels (mean=30.7, SEM=2) decreased and IL-6 levels (mean=4.4, SEM=1) increased with age as did marrow fat (mean=1.2mmfat/g, SEM=0.04). There were no significant correlations between cytokine levels or fat and SP-HSC numbers. Stem cells appear to be progressively lost with aging and only the highest quality stem cells survive.

Glomerular structural and functional changes in a high-fat diet mouse model of early-stage Type 2 diabetes

AIMS/HYPOTHESIS

Type 2 diabetes often results in diabetic nephropathy, which is preceded by an elevated glomerular filtration rate (GFR). This study was designed to develop a mouse model of Type 2 diabetes and to elucidate the glomerular events in the early stages of diabetic nephropathy.

METHODS

Four-week-old mice were fed a normal or high-fat (42% of total calories from fat) diet, and body weight, blood glucose, insulin, leptin, lipids and GFR were monitored from 9 to 21 weeks or longer after the feeding programme. Mesangial cell dedifferentiation was accessed by alpha-smooth muscle actin staining. Glomerular hypertrophy was determined using image analysis with haematoxylin-eosin staining. Matrix deposition was determined by type IV collagen staining.

RESULTS

After 9 weeks, mice fed a high-fat diet weighed more than mice fed a normal diet (30.5+/-1.2 vs 22.3+/-0.5 g, p<0.05), and mice fed a high-fat diet were hyperinsulinaemic (283.9+/-69.7 vs 102.9+/-36.4 pmol/l, p<0.05), hyperglycaemic (8.0+/-0.6 vs 6.5+/-0.2 mmol/l, p<0.05) and their leptin levels were increased six-fold (1.48+/-0.45 vs 0.25+/-0.03 ng/ml, p<0.05). After 13 weeks, mice fed a high-fat diet showed hyperfiltration (GFR; 440+/-60 vs 210+/-10 microl/min, p<0.05). During the early stages of diabetic nephropathy, mesangial cell dedifferentiation was evident, shown by increased expression of alpha-smooth muscle actin in the glomeruli. After 9 weeks, mice fed a high-fat diet already demonstrated increased type IV collagen deposition. After 13 weeks, they developed enlarged glomerular tufts compared with those of their age-matched controls.

CONCLUSIONS/INTERPRETATION

The results of this study suggest that collagen IV deposition precedes the hyperfiltration and enlargement of glomeruli in early-stage diabetic nephropathy. Dedifferentiation of mesangial cells may be associated with collagen IV deposition.

Acute effects of insulin on aqueous humor flow in patients with type 1 diabetes

PURPOSE

Previous studies reported reduced aqueous humor flow through the anterior segment of the eye in patients with type 1 diabetes. This study investigates whether reduced flow is the result of the diabetic state or of alterations in glucose or insulin concentrations.

METHODS

A cross-sectional study, involving patients with type 1 diabetes and healthy controls, measured aqueous flow at different insulin concentrations. Eleven patients with type 1 diabetes (hemoglobin A1C = 7.0 +/- 0.3% [mean +/- SEM], normal < 6.5) with no microvascular complications and 17 controls were prospectively studied. Controls were studied fasting and during a hyperinsulinemic-euglycemic clamp (insulin 2 mU/kg per minute). Patients with type 1 diabetes were similarly studied during two euglycemic clamp procedures (insulin 0.5 and 2.0 mU/kg per minute). Aqueous flow was measured by fluorophotometry. Pulsatile ocular blood flow and intraocular pressure were measured with a Langham flow probe.

RESULTS

Control subjects had no change in aqueous flow during fasting and hyperinsulinemic conditions (3.0 +/- 0.1 vs 2.8 +/- 0.1 microl per minute). In the patients with type 1 diabetes, aqueous flow was not decreased with hyperinsulinemia, compared with the low insulin state (P =.7). Compared with control subjects, patients with type 1 diabetes had lower aqueous flow during hyperinsulinemia (2.4 +/- 0.1 microl per minute, P =.03) and at lower insulin conditions (2.6 +/- 0.1 microl per minute, P <.05). No differences in intraocular pressure or pulsatile ocular blood flow were noted between groups or between insulin states within groups.

CONCLUSIONS

Aqueous flow is decreased in patients with type 1 diabetes under euglycemic conditions of high and relatively low insulin concentrations, despite the absence of microvascular complications.

Cyclosporine challenge test revisited: does it predict outcome after solitary pancreas transplantation?

BACKGROUND

The selection of patients for solitary pancreas transplantation (PTA) requires identification of individuals who will not develop acute renal dysfunction in response to immunosuppressants. A cyclosporine challenge test (CCT) was developed to predict post-PTA kidney dysfunction secondary to calcineurin inhibitor immunosuppressants. We now report on the long-term follow-up of patients who received a PTA after undergoing a CCT.

METHODS

Twelve potential PTA recipients were administered cyclosporine A (CsA) for 6 wk. Creatinine clearance (CrCl) was measured at 2, 4, and 6 wk. Those who did not fail the CCT received PTA. Baseline and post-transplant CrCl were retrospectively evaluated in the original cohort and in a group of matched patients who received PTA without a CCT.

RESULTS

Of the original 12 recipients evaluated with the CCT, 6 received PTA. CrCl was followed for a mean of 45.8 months. Of the 4 who remained alive, 2 went on to develop renal failure (CrCl < 30 mL/min) at 18 and 65 months post-transplant. The baseline CrCl was higher in PTA recipients who had not been selected to be studied with CCT than those that were (117 +/- 32 vs 78 +/- 13 mL/min). By 12 months post-PTA, the CrCl was no longer different between the groups selected to be screened with CCT and those that were not.

CONCLUSIONS

CCT may help predict risk for short-term changes in renal function (< 18 months) in response to CsA. CCT may be most helpful in candidates for PTA with borderline renal insufficiency (60-80 mL/min).

Effect of 24 hours of starvation on plasma glucose and insulin concentrations in subjects with untreated non-insulin-dependent diabetes mellitus

Adherence to a low-calorie diet often results in a decrease in blood glucose concentration in persons with non-insulin-dependent diabetes mellitus (NIDDM). Whether this is due to the resultant weight loss or to a decrease in caloric intake has been uncertain. We have obtained data previously that indicated a very short-term reduction in caloric intake (5 hours) resulted in a significant decrease in plasma glucose concentration in subjects with NIDDM. The purpose of the present study was to determine if a further decrease in glucose would occur if the fast was extended from 5 to 24 hours. Seven male subjects with untreated NIDDM were studied after an 11-hour overnight fast. For the subsequent 24-hour period, subjects were given only water. Blood was obtained for glucose, insulin, C-peptide, triglycerides, nonesterified fatty acids (NEFA) alpha-amino acid nitrogen, urea nitrogen, and glucagon at hourly intervals for 24 hours beginning at 8 AM. The amount of glycogen degraded was calculated based on the potassium balance. Plasma glucose decreased from 158 mg/dL at 8 AM to a nadir of 104 mg/dL at 7 PM. It then increased by 30 mg/dL. Corresponding changes occurred in insulin and C-peptide. Serum glucagon remained unchanged. Serum alpha-amino acid nitrogen and urea nitrogen decreased. Triglycerides and NEFA increased. The calculated glycogen utilized over this period was approximately 167 g. This would provide approximately 700 kcal energy. The elevated blood glucose concentration in mild to moderately severe untreated NIDDM subjects was normalized following short-term fasting. Plasma insulin concentrations also decreased to within normal limits. These decreases were highly significant. Glycogenolysis is an important source of fuel during this period.

Accelerated cholesteryl ester transfer in noninsulin-dependent diabetes mellitus

Alterations in core lipid composition of lipoproteins in noninsulin-dependent diabetes mellitus (NIDDM) patients have suggested that the heteroexchange of neutral lipids between HDL and the apo B-containing lipoproteins may be enhanced. For this reason, we studied cholesteryl ester transfer (CET) in ten sulfonylurea-treated patients with stable NIDDM. CET measured in all NIDDM subjects with an assay of mass transfer was significantly greater than that of controls at 1 and 2 h (P < 0.001); the transfer of radiolabeled CE also was increased in a subset of four of the NIDDM group (NIDDM k = 0.21 +/- 0.04 vs. control k = 0.10 +/- 0.05; P < 0.05). A weak correlation was demonstrable between the mass of CE transferred at 1 h and diabetic control expressed as plasma fructosamine (r = 0.58, P < 0.09). To characterize this disturbance in CET further, the donor (HDL + VHDL) and acceptor (VLDL + LDL) lipoprotein fractions were isolated by ultracentrifugation at d 1.063 g/ml from NIDDM and control plasma and a series of recombination experiments were performed. Combining NIDDM acceptor with control donor fractions that contained HDL and CETP and not the combination of NIDDM donor and control acceptor lipoproteins resulted in an accelerated CET response identical to that observed in NIDDM whole plasma. This observation indicated that the abnormality in CET in NIDDM was associated with the VLDL + LDL fraction.(ABSTRACT TRUNCATED AT 250 WORDS)

Plasma glucose and insulin responses to bananas of varying ripeness in persons with noninsulin-dependent diabetes mellitus

With progressive ripeness there is a decrease in starch and an increase in free sugar content of bananas. The starch also is considered to be poorly digestible. Therefore, we decided to study plasma glucose, serum insulin, C-peptide, and plasma glucagon responses to bananas with increasing degrees of ripeness. Seven male subjects with untreated noninsulin-dependent diabetes mellitus ingested 50 g carbohydrate as bananas of stage 4 (more yellow than green), 5 (yellow with green tip), 6 (all yellow), and 7 (yellow flecked with brown) ripeness. They also received 50 glucose on two occasions for comparative purposes. On a separate occasion water only was given as a control. The area responses were quantified by determining incremental areas using the water control as baseline. The mean glucose area following the 50 g glucose meals was 15.1 +/- 1.9 mM.h. After the ingestion of bananas of 4, 5, 6 and 7 ripeness the glucose area response was 42, 41, 51 and 48% of that after glucose ingestion, respectively. The insulin area response following glucose meals was 888 pM.h. Responses to 4, 5, 6 and 7 bananas were 85, 70, 61, 85%, respectively, of that following glucose ingestion. C-peptide data were similar to the insulin data. The glucagon area response was negative after glucose ingestion but was positive following banana ingestion. In summary, the glucose, insulin, C-peptide, and glucagon area responses varied little with ripeness of the bananas.

Metabolic response to cottage cheese or egg white protein, with or without glucose, in type II diabetic subjects

Test meals with 25 g protein in the form of cottage cheese or egg white were given with or without 50 g glucose to male subjects with mild to moderately severe, untreated, type II diabetes. Water was given as a control meal. The glucose, insulin, C-peptide, alpha amino nitrogen (AAN), glucagon, plasma urea nitrogen (PUN), nonesterified fatty acid (NEFA), and triglyceride area responses were determined using the water meal as a baseline. The glucose area responses following ingestion of cottage cheese or egg white were very small compared with those of the glucose meal, and were not significantly different from one another. The serum insulin area response was 3.6-fold greater following ingestion of cottage cheese compared with egg white (309 v 86 pmol/L.h). The simultaneous ingestion of glucose with cottage cheese or egg white protein decreased the glucose area response to glucose by 11% and 20%, respectively. When either protein was ingested with glucose, the insulin area response was greater than the sum of the individual responses, indicating a synergistic effect (glucose alone, 732 pmol/L.h; glucose with cottage cheese, 1,637 pmol/L.h; glucose with egg white, 1,213 pmol/L.h). The C-peptide area response was similar to the insulin area response. The AAN area response was approximately twofold greater following ingestion of cottage cheese compared with egg white. Following ingestion of glucose, it was negative. When protein was ingested with glucose, the AAN area responses were additive. The glucagon area response was similar following ingestion of cottage cheese or egg white protein. Following glucose ingestion, the glucagon area response was negative.(ABSTRACT TRUNCATED AT 250 WORDS)

The metabolic response to various doses of fructose in type II diabetic subjects

Eight men with untreated type II diabetes were given 480 mL water containing 15 g, 25 g, 35 g, and 50 g fructose orally, in random sequence. The same subjects were given the same volume of water as a control. They also were given 50 g glucose on two occasions for comparative purposes. Plasma glucose, urea nitrogen, and glucagon, and serum insulin, C-peptide, alpha-amino-nitrogen (AAN), nonesterified fatty acids (NEFA), and triglycerides were determined over the subsequent 5-hour period. The area responses to each dose of fructose were calculated and compared with the water control. The integrated glucose area dose-response was curvilinear, with little increase in glucose until 50 g fructose was ingested. With the 50-g dose, the area response was 25% of the response to 50 g glucose. The insulin response also was curvilinear, but the curve was opposite to that of the glucose curve. Even the smallest dose of fructose resulted in a relatively large increase in insulin, and a near-maximal response occurred with 35 g. The area response to 50 g fructose was 39% of that to 50 g glucose. The C-peptide data were similar to the insulin data. The AAN area response to fructose ingestion was negative. However, the response was progressively less negative with increasing doses. The glucagon area response was positive, but a dose-response relationship was not apparent. The glucagon area response was negative after glucose ingestion, as expected. The urea nitrogen area response was negative, but again, a dose-response relationship to fructose ingestion was not present.(ABSTRACT TRUNCATED AT 250 WORDS)

Prolonged fasting reduces rat hepatic beta 1 thyroid hormone receptor protein without changing the level of its messenger ribonucleic acid

The level of hepatic nuclear T3-binding capacity falls in rats subjected to fasting. To define the mechanism underlying these changes, we have assayed in liver the concentration of the mRNA coding for the beta 1-receptor (beta 1-TR) isoform, the total nuclear T3-binding capacity, and the fraction of the total binding capacity that can be specifically immunoprecipitated with an anti-beta 1-TR immunoglobulin G preparation. Although no changes in beta 1-TR mRNA concentration were noted, we observed a 60% fall in total binding capacity. beta 1-TR mRNA levels were preserved despite a 50% fall in total poly(A)+ RNA. The fall in beta 1-TR protein, however, was consistent with a generalized decrease in total hepatic protein content. This study provides yet another instance in which measurement of receptor mRNA is not consonant with the behavior of the nuclear T3 receptor protein.

Lipoprotein composition and HDL particle size distribution in women with non-insulin-dependent diabetes mellitus and the effects of probucol treatment

To further characterize the spectrum of potentially atherogenic disturbances in lipoprotein composition in non-insulin-dependent diabetes mellitus (NIDDM), we have studied a subset of women with NIDDM before and after treatment with the lipophilic lipid-lowering drug probucol (1 gm day), which we have shown corrects certain compositional abnormalities these women share with subjects who have hypercholesterolemia. Before treatment, the NIDDM group had a somewhat higher plasma triglyceride level (154 +/- 58.3 mg/dl, vs control, 80.0 +/- 21 mg/dl [mean +/- SD]; p less than 0.025) than controls but their cholesterol and high-density lipoprotein cholesterol (HDL-C) levels did not differ from control levels. A number of significant disturbances, however, were present in the surface and core lipid composition of their lipoproteins. Although the cholesterol content of NIDDM low-density lipoprotein (LDL) was similar to that of controls, its content of sphingomyelin and phosphatidylinositol plus phosphatidylserine and sphingomyelin-to-lecithin ratio all were significantly reduced. Moreover, their very-low-density lipoprotein (VLDL) and HDL2 tended to have reduced amounts of free (unesterified) cholesterol (FC) relative to lecithin, and their HDL2 and HDL3 tended to be triglyceride enriched. Probucol therapy resulted in significant decreases in total plasma cholesterol (-15%), FC (-28%), HDL-C (-22%), and triglyceride (-16%) and in apoproteins A-I, B, and E (apo A-I, B, and E), without changing diabetic control (before probucol: hemoglobin A1, cholesterol, 10.7% +/- 2.7%; after probucol: 10.9% +/- 3.0%).(ABSTRACT TRUNCATED AT 250 WORDS)

Functional relationship of thyroid hormone-induced lipogenesis, lipolysis, and thermogenesis in the rat

Metabolic balance studies were carried out to determine the interrelationships of thyroid hormone-induced lipogenesis, lipolysis, and energy balance in the free-living rat. Intraperitoneal doses of 15 micrograms triiodothyronine (T3)/100 g body wt per d caused an increase in caloric intake from 26.5 +/- 1.7 (mean +/- SEM) kcal/100 g per d to 38.1 +/- 1.5 kcal/100 g per d. Food intake, however, rose only after 4-6 d of treatment and was maximal by the 8th day. In contrast, total body basal oxygen consumption rose by 24 h and reached a maximum by 4 d. Since total urinary nitrogen excretion and hepatic phosphoenolpyruvate carboxykinase mRNA did not rise, gluconeogenesis from protein sources did not supply the needed substrate for the early increase in calorigenesis. Total body fat stores fell approximately 50% by the 6th day of treatment and could account for the entire increase in caloric expenditure during the initial period of T3 treatment. Total body lipogenesis increased within 1 d and reached a plateau 4-5 d after the start of T3 treatment. 15-19% of the increased caloric intake was channeled through lipogenesis, assuming glucose to be the sole substrate for lipogenesis. The metabolic cost of the increased lipogenesis, however, accounted for only 3-4% of the T3-induced increase in calorigenesis. These results suggest that fatty acids derived from adipose tissue are the primary source of substrate for thyroid hormone-induced calorigenesis and that the early increase in lipogenesis serves simply to maintain fat stores. Since the mRNAs coding for lipogenic enzymes rise many hours before oxygen consumption and lipolysis, these results suggest that T3 acts at least in part by an early coordinate induction of the genes responsible for these processes.

Persistent abnormalities in lipoprotein composition and cholesteryl ester transfer following lovastatin treatment

Optimally effective lipid-lowering agents should not only restore plasma lipids to normal levels but also correct potentially atherogenic alterations in lipoprotein composition and function often present in hyperlipidemic patients. Lovastatin, a competitive inhibitor of cholesterol biosynthesis, clearly lowers plasma cholesterol levels. Its effects on lipoprotein composition and cholesteryl ester transfer (CET), a key step in reverse cholesterol transport, however, are not known. Since abnormalities in CET and lipoprotein composition are present in patients with hypercholesterolemia, we studied these parameters of plasma lipoprotein transport in twelve hypercholesterolemic (HC; Type IIa) subjects (six male, six female) before and 2 months after lovastatin treatment (20 mg qd). Before lovastatin, the free cholesterol (FC)/lecithin (L) ratio in plasma, a new index of cardiovascular risk that reflects lipoprotein surface composition, was abnormally increased (1.18 +/- 0.26 vs controls 0.83 +/- 0.14; P less than 0.001) in very low density lipoproteins (VLDL) and high density lipoprotein-3 (HDL3), and remained so after treatment despite significant declines in whole plasma cholesterol (311.7 +/- 68.2 vs 215.6 +/- 27.2 mg/dl; P less than 0.001), low density lipoprotein (LDL)-cholesterol (206.3 +/- 47.9 vs 146.8 +/- 29.4; P less than 0.001), and apolipoprotein B (149 +/- 30 vs 110 +/- 17; P less than 0.005).(ABSTRACT TRUNCATED AT 250 WORDS)

Immune abnormalities associated with HLA-B8: lymphocyte subsets and functional correlates

Mononuclear cell populations were enumerated in healthy young adults with or without the histocompatibility antigen HLA-B8. Mononuclear cell counts were lower in subjects with HLA-B8, as was reflected in lower absolute numbers of the cell subsets. When cell populations were compared as percentages of total mononuclear cells, subjects with HLA-B8 had significantly more B lymphocytes bearing IgM than did subjects without HLA-B8. The T4/T8 ratio was significantly increased in subjects with HLA-B8. due both to increases in OKT4+ cells and to decrease in OKT8+ cells. B-Lymphocyte function was compared in subjects with and without HLA-B8 by measuring pokeweed mitogen driven differentiation of B cells to IgM bearing blasts, and was significantly greater in subjects with HLA-B8. Subjects with HLA-B8. Subjects with HLA-B8 were also found to have decreased Con A-induced suppression. Alterations in lymphocyte subsets and lymphocyte functions may underlie the predisposition to autoimmunity associated with HLA-B8.

The effects of complement activation by cobra venom factor on the migration of T and B lymphocytes into rat thoracic duct lymph

Experiments were done to see whether C3 or C3-split products are involved in lymphocyte recirculation, with particular reference to B lymphocytes which have C3b receptors. Rats were injected with cobra venom factor (CVF), and the output of subclasses of lymphocytes was measured in thoracic duct lymph in hourly collections during the subsequent 24 h. During the period of acute C3 activation which lasted for 2-8 h, the output of lymphocytes decreased by 47%, but returned to normal at later times, when C3 levels were reduced to less than 20% normal. There was no effect on the output of C3b receptor lymphocytes, and this receptor was not blocked probably because initial C3 levels in lymph were only 13% of blood levels, so that only small amounts of C3b were generated in lymph. When these lymphocytes were labelled and injected i.v. they migrated with the slow rate which is characteristic of normal B lymphocytes. The main effect of CVF was to reduce the output of T lymphocytes by 58% during the phase of acute C3 activation. When normal thoracic duct lymphocytes were labelled and injected, their rate of reappearance in thoracic duct lymph was only reduced during this phase. It was concluded that recirculation of lymphocytes is not C3 dependent, and that insufficient C3b is generated in lymphoid tissues to block C3b receptors on B lymphocytes during periods of rapid C3 activation. However the migratory rate of T lymphocytes through these tissues is reduced during this period, and it is suggested that this may be due to an effect of C3 split products on macrophages which lie along T-lymphocyte traffic routes.

Lymphocyte stimulation with skin cells. A function of maturation

Cell monolayers grown from trypsin dispersions of whole rat embryos were tested for their capacity to stimulate allogeneic rat thymus cells in mixed cell cultures. Cells from whole embryos of only 15-days gestation were virtually ineffective at stimulating allogeneic thymus cells. Cells from 19-day-old whole embryos produced definite stimulation and stronger stimulation was obtained with skin cell monolayers prepared from these embryos. The average amount of allogeneic stimulation obtained with rat skin cells increased with the age of the donor animals. The greatest degree of stimulation occurred when the skin cells were obtained from rats about 1 week postpartum. Thereafter, the stimulating potential of allogeneic skin cells declined and was found to stabilize at the levels observed when adult skin cells were used as the stimulator cells. Neither rat brain cells nor rat kidney cells (taken from embryos, neonates, and adults) were capable of inducing allogeneic rat thymocytes to undergo significant transformation in culture. We conclude that the in vitro lymphocyte response to skin cell stimulation is a function of skin cell maturation, and that the inability of kidney cells or brain cells to provoke the same sort of stimulation is not attributable to antigen loss during or after foetal development, but exists as a feature of these cells at the earliest stages of differentiation.