The role of beta-cell death and regeneration
The loss of β-cells in T2DM has been mainly attributed to increased β-cell death due to apoptosis and other forms of cell death [6,9,13,46 – 49], possibly driven by adverse environmental conditions [50 – 52] and probably mediated by several intracellular mechanisms [50 – 54]. Apoptosis is a type of programmed cell death morphologically characterized by cell rounding up, bleb formation and chromatin condensation (Figures 24.2(a) and 24.2(b)). As a matter of fact, in autoptic pancreatic samples apoptosis has been shown to be significantly increased in both obese and lean type 2 diabetic cases as compared to BMI-matched, nondiabetic controls . In another study, β-cell apoptosis was found to be threefold increased in obese type 2 diabetic patients , although the diabetic condition did not affect apoptosis in lean individuals. Increased β-cell apoptosis in diabetic islets has been confirmed by electron microscopy . In addition, by assessing cytoplasmic histone-associated DNA fragments, it has been observed that islet cell death is significantly enhanced in isolated diabetic islets . These changes are accompanied by increased numbers of cells positive for activated caspase-3  as well as greater activity of caspase-3 and caspase-8 , which are key molecules in the induction and execution of apoptosis.
However, forms of programmed cell death other than apoptosis have been described . One involves autophagy, which is a type of cell death that occurs without marked chromatin condensation and is accompanied by massive vacuolization of the cytoplasm [55 – 57]. In general, normally functioning autophagy has a beneficial role for cells (including the β-cell [58 – 60]) as it regulates the turnover of aged proteins and eliminates damaged structures and organelles . However, cells that undergo altered autophagy may die in a nonapoptotic manner [55 – 57]. The presence of autophagy was investigated in β-cells from T2DM and matched nondiabetic subjects . On electron microscopy, there were significantly more dead β-cells in diabetic than control samples; while several of these cells had morphologic signs of apoptosis, massive vacuole overload (suggesting altered autophagy) was associated with a proportion of dead β-cells without signs of apoptosis (Figure 24.2(c)). This proportion was significantly higher in type 2 diabetic samples. It can be therefore concluded that β-cell death is increased in type 2 diabetic patients due to enhanced apoptosis and other forms of cell death, which may contribute to β-cell failure in this disease.
Whether defects of β-cell regeneration also plays a role in the reduction of β-cells in human T2DM is still unclear. β-cell regeneration may essentially occur by replication (proliferation) of existing cells, neogenesis from precursors or transdifferentiation of existing mature cells [61 – 67]. Normally, insulin-positive cells appear in the human pancreas at around the 8th week of gestation; at 10 weeks postconception all clusters containing more than 10 insulin-positive cells have developed a close relationship with vascular structures . After an additional 2 – 3 weeks, human fetal islets contain cells independently immunoreactive for insulin, glucagon, somatostatin and pancreatic polypeptide . During prenatal pancreas development, there is a linear increase in fractional β-cell area, which reaches a value of ∼3% at birth . β-cell proliferation is very efficient during fetal life, involving more than 3% of β-cells during week 17 – 32 of gestation [69,70]; this proportion then tends to decrease and approximates 1.5 – 2.0% at birth [69,70]. After birth, as shown by the study of autoptic pancreatic samples from subjects aged 2 to 20 years, β-cell mass expands several-fold from infancy to adulthood, mainly due to growth in islet size rather than number, a process driven by the rate of β-cell replication . After the age of 20–30 years, however, the rate of islet β-cell replication seems β-cell complement is fully established by the age of 20 – 30 years. This implies that β-cell regeneration in adults may occur at a very low rate, if at all.
Nevertheless, β-cell mass in adult human individuals can increase, such as occurs in obesity and during pregnancy [5 – 7,9,73,74]. In one of the studies mentioned earlier , it was observed that nondiabetic obese subjects (with an average BMI of ∼35 kg m−2 ) show a 50% increase in β-cell volume as compared to nondiabetic lean subjects (average BMI of ∼23 kg m−2 ). Apoptosis and replication did not differ significantly between obese and nonobese cases, and the increase of β-cells was attributed to enhanced neogenesis, as indirectly suggested by the greater number of insulin positive cells in or close to the ducts . However, the same authors, in a more focused article, while confirming the augmented β-cell mass in autoptic pancreatic samples of obese subjects, were unable to find differences in terms of apoptosis or regeneration, as assessed by some of the currently available surrogate markers (Ki67, insulin-positive cells in the duct wall) . In another report , obesity (average BMI of 31 kg m−2 ) was associated with a twofold increase in β-cell volume compared to lean (aver- age BMI of 24 kg m−2 ) nondiabetic individuals. In this case, the authors found that both β-cell replication and neogenesis were significantly enhanced in the obese samples in the face of similar rates of apoptosis.
Pregnancy is also associated with greater β-cell mass . Although, for obvious reasons, data in humans are very scanty, one study has reported on the morphometry of the pancreatic islets during gestation in humans . The authors collected pancreases obtained at autopsy from women who had died while pregnant, and found that the pancreatic fractional β-cell area was increased by approximately 1.4-fold compared to nonpregnant women . Mean β-cell size was not different, and in pregnancy there were more small islets rather than an increase in islet size or β-cells per islet. No increase in β-cell negligible, as shown in two independent studies analyzing lipofuscin accumulation and thymidine incorporation in human β-cells [71,72]. Both these studies concluded that, in human islets, the replication or change in β-cell apoptosis was detected, but duct cells positive for insulin and scattered β-cells were increased with pregnancy, again suggesting, although indirectly, a possible preminent role of β-cell neogenesis. Interestingly, in a case of gestational diabetes it has been observed that total insulin area was reduced, not due to increased apoptosis but reduced regeneration .
The issue of β-cell regeneration has been investigated in a few morphometric studies with pancreases from nondiabetic and type 2 diabetic individuals (Table 24.2). It was initially reported that there was no significant difference in the frequency of β-cell replication (as assessed by Ki67 protein staining) between obese nondiabetic and type 2 diabetic subjects or lean nondiabetic and type 2 diabetic subjects . Similarly, when neogenesis was indirectly quantified by counting duct cells immunoreactive for insulin, no difference was found between the obese or lean nondiabetic and type 2 diabetic cases . However, in another report slightly different results were obtained . When β-cell replication was examined by proliferating cell nuclear antigen (PCNA) staining, this was lower in obese type 2 diabetic than in weight-matched nondiabetic cases; on the other hand, a “neogenic” index calculated from the amount of insulin positive cells or small clusters in the duct walls and/or the acinar tissue was found to be similarly increased in nondiabetic obese individuals and in obese or lean type 2 diabetic patients as compared to nondiabetic lean subjects . Three additional studies were published thereafter [12,14,46], all showing no difference between diabetic and nondiabetic samples in the number of cells co-stained for insulin and Ki67; however, in one report the prevalence of insulin-positive cells in the duct walls was found to be significantly higher in subjects with impaired glucose tolerance and significantly lower in those with T2DM, compared to nondiabetic cases .
Altogether, the available information, obtained with the use of surrogate markers and indexes of regeneration, indicates that β-cell replication does not seem to differ between type 2 diabetic and nondiabetic β-cells, whereas neogenesis may or may not be reduced in T2DM. Clearly, even if present, in T2DM β-cell regeneration is not sufficient to replace the β-cells that have been demised.