Beta-Cell response to intravenous glucose
Although in normal living conditions β cells are stimulated by hyperglycemia that follows glucose ingestion, the study of the response to intravenous glucose is of fundamental importance for understanding the physiology of β cells. Several tests have been developed for this purpose and this section describes the most relevant and the characteristics of insulin secretion that they reveal.
The hyperglycemic clamp
The most typical test for the study of β-cell response employs a brisk and sustained elevation of glucose concentration from a baseline value. In vivo, this is typically achieved with the hyperglycemic clamp technique (Figure 8.2) . The β-cell response to this glucose stimulus is biphasic, with an initial insulin secretion burst lasting about 5–8min (first-phase secretion), followed by a drop towards basal levels and then by a relatively rapid increase that persists as far as hyperglycemia is maintained (second-phase secretion).
One important physiologic feature highlighted by the hyperglycemic clamp is that the β cell responds to the increase in glucose in a proportionate manner. In other words, β-cell function, as for many other physiologic processes, can be described through a dose-response curve describing the relationship between insulin secretion and glucose concentration in steady-state conditions (Figure 8.2). This dose-response can be assessed with greater accuracy using a stepped increase in glucose concentration, the so-called graded glucose infusion test . In this test, glucose is infused intravenously at increasing rates and glucose and C-peptide are measured. Insulin secretion is calculated by deconvolution of C-peptide and the β-cell dose-response is constructed by plotting insulin secretion rates at the end of each glucose infusion period versus the corresponding glucose concentration.
The β-cell dose-response is a fundamental characteristic of insulin secretion; its slope, which represents the sensitivity of the β cell to glucose, is a key β-cell function parameter. In healthy subjects insulin secretion at a basal glucose level of 5 mmol L−1 is about 90 pmol min−1 and increases by five- to sixfold (500 pmol min−1) at 10 mmol L−1 . The relationship is fairly linear within this glucose concentration range and above; under these circumstances, β-cell glucose sensitivity (i.e., the dose-response slope) is about 80 pmol min−1 per mmol L−1.
First-phase secretion becomes evident with a brisk elevation of glucose concentration.The amount of insulin secreted during the first phase is dependent on the magnitude of the glucose increase, so that even the first phase recognizes a dose-response feature ; in a typical +7 mmol L−1 hyperglycemic clamp it is around 4 nmolm−2 representing the 10–15% of what is secreted per hour in the second phase . Though limited, the amount of insulin secreted in the first phase is relevant at least for two reasons. First, while the secretion burst is apparent with a rapid elevation in glucose concentration, the underlying secretory mechanisms appear to be active even for a more gradual rise in glucose concentration . These mechanisms are likely to be responsible for a response that is anticipated compared to what would be predicted solely on the basis of the dose-response and this anticipation has relevant physiologic implications for glucose homeostasis, as discussed later. Second, first-phase insulin secretion is a very sensitive marker of early β-cell dysfunction . Impairment of first-phase secretion is already present in subjects at risk of developing diabetes , and is predictive of diabetes onset . Because of these relevant characteristics, assessment of first-phase secretion has been widely used.
The intravenous glucose tolerance test
A common technique for the assessment of first-phase insulin secretion is the intravenous glucose tolerance test (IVGTT, also known as FSIGTT, frequently sampled intravenous glucose tolerance test), in which glucose (typically 0.3 g kg−1) is injected intravenously over a short period of time (2–3 min). The sharp rise in glucose concentration elicits a first-phase response similar to that observed in the initial part of the hyperglycemic clamp, which is commonly quantified using the so-called acute insulin response (AIR) . This widely used index is typically calculated as the mean increment above baseline in insulin concentration in the first 8–10 min. An advantage of the IVGTT is its relative experimental simplicity, compared to the clamp. Moreover, similarly to the hyperglycemic clamp, parameters of insulin sensitivity can be derived as well . Many of the studies concerning the relationships between insulin sensitivity and insulin secretion, discussed later, are based on this approach . In contrast to AIR, late-phase insulin secretion is not commonly calculated during IVGTT testing.
When evaluated using C-peptide deconvolution, the amount of insulin secreted during the IVGTT first phase is estimated to be ∼3 nmolm−2, a value similar to that of the hyperglycemic clamp [30,31].
In summary, insulin secretion when assessed with intravenous glucose challenges appears to be highly dynamic with two main characteristic phases, a rapid and transient one followed by a more sustained one. These responses can be reproduced in vitro, by pancreas perfusion as well as in perifused islets [9,32]. Nonetheless, the mechanisms underlying the biphasic response remain only partially understood. It has been proposed that first-phase secretion is the consequence of the discharge of a pool of insulin granules located in the proximity of the cell membrane in response to an increase in intracellular calcium triggered by a cascade of electrochemical events generated by glucose utilization inside the β cell . Exocytosis, however, is very sensitive to calcium levels and the typical early peaking of intracellular calcium  may well exert a direct contribution to this phenomenon. Second-phase secretion is also controlled by calcium, but glucose itself, independently from calcium, plays a role in sustained insulin secretion .
Beta-Cell response to oral glucose
The amount of insulin secreted above the basal levels during a 2-h 75 g OGTT in normal subjects is ∼30 nmolm−2 . The insulin response to glucose ingestion is more pronounced than that elicited by an intravenous infusion achieving the same glucose levels. This difference becomes readily apparent when the plasma glucose profile after an OGTT is reproduced by means of an intravenous infusion of glucose [36,37]. The greater response to oral glucose is referred to as the “incretin effect.”
The incretin effect
As shown in Figure 8.3, when glucose concentration changes during the OGTT, insulin secretion follows the glucose pattern, although insulin secretion is considerably higher after oral ingestion. This implies that the β-cell dose-response for oral glucose is shifted upwards compared to that for intravenous glucose (Figure 8.3). Estimates of the magnitude of the potentiating effect of oral glucose in normal subjects vary from study to study, partly because of the different methods used to calculate the incretin effect. Recent studies based on modeling analysis report an increase in insulin secretion of ∼1.6–1.7-fold with oral (75 g OGTT) compared to intravenous glucose administration [38,39]. The incretin effect appears to be quite variable also within subjects with normal glucose tolerance, ranging from a negligible effect to a two- to threefold amplification. The incretin effect is dose-dependent, that is, higher glucose doses elicit stronger effects; for a fivefold increase in the glucose load, the increase in incretin effect is almost twofold, from ∼1.2 to ∼2.4 [38,40].
The incretin effect is mainly attributed to the action of two hormones: glucose-dependent insulinotropic peptide (GIP) and glucagon-like peptide 1 (GLP-1) . GIP and GLP-1 are secreted as glucose and other nutrients reach sections of the intestine where cells specialized in the secretion of these hormones are located (the K cells for GIP and the L cells for GLP-1) [42,43]. GIP and GLP-1 bind to specific receptors on the β cell that activate a cascade of events leading to increased insulin secretion. The action of these hormones is glucose-dependent, that is, the increase in insulin secretion is higher when glucose concentration is higher, whereas at low glucose the amplification is marginal. GIP and GLP-1 may also act indirectly on the β cell through stimulation of the nervous systems, as receptors for these hormones are present in neuronal cells . The secretion of GIP and GLP-1 is dependent on the amount of nutrients reaching the intestine.
Experiments employing exogenous infusion of GIP and GLP-1 have clearly shown that these hormones stimulate insulin secretion in a concentration-dependent fashion [45,46]. In particular, GLP-1 has been shown to make the β cell dose-response steeper , similarly to Figure 8.3. However, whether GIP and GLP-1 levels can entirely account for the incretin effect has not been fully established. The relationship between the levels of incretin hormones and the incretin effect are in fact generally weak [38,47,48]; whether this is due to limited assay precision or factors other than the hormone levels remains to be clarified.
Insulin secretion mechanisms during an oral glucose load
OGTT modeling analysis has enabled more careful assessment of the multiple components of insulin secretion (Figure 8.4) . According to this approach, the β cell dose-response is responsible for most of the changes in insulin secretion during an OGTT. Early secretion mechanisms, which enhance insulin secretion during the initial part of the OGTT, contribute only one tenth (∼3 nmolm−2) of total supra basal secretion (∼30 nmolm−2) (unpublished results from ). Notably, the magnitude of the early insulin secretion component is similar to that calculated during the hyperglycemic clamp or an IVGTT.Therefore, based on this analysis the contribution of early secretion phenomena is quantitatively limited during an oral glucose load as well.
The response to oral glucose is also characterized by phenomena that enhance β-cell glucose sensitivity during the test, referred to as “potentiation.”This mechanism may well account for persistent activation of insulin secretion at the end of the OGTT when glucose concentration has returned to baseline levels [20,31]. Modeling analysis predicts that in normal subjects at the end of a 2-h OGTT insulin secretion is ∼70% higher compared to baseline at the same glucose levels .
As described earlier (Figure 8.3), a key effect of the stimulation of the incretin system is the upward shift of the β-cell dose-response, that is, an increase in glucose sensitivity. In addition, the incretin effect also enhances early secretion [38,39]. The influence of the incretin effect on potentiation is more complex. A recent study  has shown that the potentiation phenomena observed during an OGTT are a combination of the incretin effect and glucose-induced potentiation, a mechanism described later in this chapter.
Insulin secretion in normal living conditions
During a 24-h observation period in which meals are administered at the usual times, insulin secretion excursions basically follow glucose excursions during the multiple meals , similar to a single OGTT. However, some studies have reported that the β-cell dose-response does not remain the same during the 24 hours. In particular, in relation to the glucose levels insulin secretion is relatively potentiated during the morning meal and attenuated during the following meals and the night, to a significant extent [20,50].