Amylin Is the More “Basal”
of the Two Beta-Cell Hormones

First, a word about clinical assays for hormones.  Peptides are measured with immunoassays which can be afflicted with specificity and calibration problems.  Insulin and glucagon have been measured since the late 1960s, and it may be difficult to directly compare results over time as assay technology has evolved.  Amylin has been measured since the early 1990s, and early assays were plagued by difficulties in working with the human amylin peptide, which tends to aggregate.  So, the results reviewed in this section may be subject to adjustments when clinical research can validate them with the latest monoclonal immunometric technology.  But these preliminary observations should be useful for defining more detailed hypotheses to be validated in the clinic.

Also, several amylin researchers have complained that, as of early 2020, there don’t appear to be reliable amylin assays on the market.  They report that amylin testing results are deemed not usable in publications.  As a result, we are unable to find recent studies that present diurnal profiles of circulating amylin in healthy nondiabetics comparable to those reporting insulin and glucagon profiles.  So, at this stage, we are reduced to presenting old, usually small studies that give, at best, a fragmentary picture.

* * * * * * *

By 1989 it was known that insulin and amylin are colocalized in the secretory granules of beta-cells. [1] Shortly thereafter it was shown that glucose ingestion stimulates increases in circulating insulin and amylin, and that there is a correlation between insulin and amylin: [2]

These observations led to the general perception that insulin and amylin are co-secreted in parallel by beta-cells, resulting in similar diurnal profiles.  An example appears in the FDA-approved SYMLIN Prescribing Information published by the manufacturer of pramlintide (Exhibit 3). [3]

Published literature generally echoes this view:

• A textbook example: “Amylin is secreted in equal proportions to insulin from beta-islet cells and causes a reduction in glycogenesis in skeletal muscle, a decrease in postprandial glucagon secretion, slows gastric emptying, and suppresses appetite.”  (2013) [4]

• A review article example: “In general, amylin secretion from the b-cell correlates very tightly with insulin secretion.” (2015)  [5] (Note that the R-squared values of 0.55 and 0.42 reported above can hardly be considered “very tightly.”)

• We have been unable to find any published sources that say the diurnal profiles of insulin and amylin differ in any important way.

But, a closer look at the chart in Exhibit 3 reveals a graphical trick:  The vertical scales of the hormone concentrations have been graphed with separate vertical axes to allow both hormone profiles to be plotted on the same chart:

• Scale of values:  The amylin scale of concentrations is about 4% of the insulin scale.  The much lower plasma concentration of amylin is consistent with its role as a neuroendocrine hormone directed at brain receptors, while insulin’s peripheral actions regulate a greater tissue mass of liver, muscle, and fat.  This scaling of the graph is legitimate for visually comparing the timing of diurnal patterns.

• Range of values:  However, while the insulin concentration axis starts at zero, the amylin axis starts at 5 picomolar, thereby forcing an alignment of the insulin and amylin diurnal profiles.  This distortion in the amylin range of values is misleading with respect to the magnitude of hormone basal and prandial levels.

When Exhibit 3 is plotted correctly, the plasma profiles diverge (Exhibit 4).

While the plasma peaks are aligned at mealtimes, the molar ratio of the two hormones varies about 4-fold over a 24-hour period, from absorptive (mealtime) ratios of over 30:1 insulin:amylin, to postabsorptive (between meals) ratios of about 8:1 (Exhibit 5).

The sample size of human subjects that generated this data is relatively small (n = 6), so further study is warranted.  However, these results are consistent with our hypothesis that, contrary to conventional wisdom, insulin and amylin diurnal profiles do NOT align precisely: while the timings of the diurnal pulses are in alignment, the ratios of resulting hormone circulating concentrations appear to vary widely.

Several other studies support this conclusion.  Exhibit 6 shows the molar insulin/amylin ratios measured for healthy, nondiabetic subjects following either mixed meals or 75 gram oral glucose tolerance tests. [6] [7] [8] [9] [10] It should be noted that these five studies were performed in the early 1990s before standardized commercial assays for amylin were available; consequently, the absolute ratios of insulin to amylin are not directly comparable.

These data demonstrate changes in insulin/amylin molar ratios of two- to four-fold following meals.  This finding supports the hypothesis that the reason for two beta-cell hormones is that the appropriate diurnal profile for regulating blood glucose influx (amylin) is different from that regulating efflux (insulin).

Quoting the conclusion in yet another study:  “We conclude that the profile of plasma total and nonglycosylated amylin concentrations, as determined under similar steady-state hyperglycemic conditions in subjects with varying degrees of glucose tolerance, differ markedly from that of insulin. These differences could reflect either slower clearance of amylin than insulin or differences in the secretory dynamics of the two peptides.” [11]

Another way to visualize the difference in beta-cell hormone profiles is by separating the areas under the curves into basal (fasting) vs. bolus (mealtime) components, as shown in Exhibits 7 and 8 (Exhibits 10 and 11 of Part 1).

These data suggest that, in healthy non-diabetics, about two-thirds of insulin’s total daily exposure is associated with mealtime peaks (boluses), whereas only about one-third of amylin’s daily exposure is bolus:

Insulin is well documented to be predominantly a bolus signal with relatively modest basal levels between meals.  This is consistent with aggressively pushing circulating glucose into liver, muscle, and fat at mealtimes, while mostly shutting down insulin-stimulated efflux between meals when brain and other tissues are consuming plasma glucose without beta-cell stimulation.  As discussed in Appendix A, R-squareds for the glucose-insulin diurnal profiles are in the 0.90+ range.

Amylin, in contrast, appears to be primarily a basal signal with relatively modest peaks at mealtime.  This diurnal pattern is consistent (1) with maintaining tonic inhibition of alpha-cells between meals as predicted by the amylin circuit-breaker model, and (2) with moderating glucose influx at mealtime by slowing gastric emptying and suppressing glucagon secretion.

If the goal of amylin replacement therapy is to properly restore amylin’s regulatory role in T1D, how well do exogenous amylin agonist infusions mimic the healthy, nondiabetic diurnal profile of endogenous amylin in healthy nondiabetics?

Endnotes:

[1]  Co-localization of islet amyloid polypeptide and insulin in the B cell secretory granules of the human pancreatic islets; Diabetologia 32:240-4 1989.

[2]  Effects of Meal Ingestion on Plasma Amylin Concentration in NIDDM and Nondiabetic Humans; Diabetes 39:752-6 1990.

[3]  24 Hour Plasma Amylin Profiles Are Elevated in IGT Subjects vs. Normal Controls; Diabetes 44(suppl.1):238A 1995.

[4]  Understanding Diabetes: A Biochemical Perspective, R. F. Dods, Summary Box 10.5 (Wiley 2013)

[5]  Amylin: pharmacology, physiology, and clinical potential; Pharmacological Reviews 67:564-600 2015.

[6]  Effects of Meal Ingestion on Plasma Amylin Concentration in NIDDM and Nondiabetic Humans; Diabetes 39:752-6 1990.

[7]   Effect of dexamethasone on insulin sensitivity, islet amyloid polypeptide and insulin secretion in humans; Diabetologia 36:84-7 1993.

[8]  Plasma islet amyloid polypeptide (Amylin) levels and their responses to oral glucose in Type 2 (non-insulin-dependent) diabetic patients; Diabetologia 34:129-32 1991.

[9]  Islet Amyloid Polypeptide Response to Glucose, Insulin, and Somatostatin Analogue Administration; Diabetes 39:639-42 1990.

[10]  Islet amyloid polypeptide plasma concentrations in individuals at increased risk of developing Type 2 (non-insulin-dependent) diabetes mellitus; Diabetologia 35:291-3 1992.

[11]  Deficiency of Total and Nonglycosylated Amylin in Plasma Characterizes Subjects with Impaired Glucose Tolerance and Type 2 Diabetes; Journal of Clinical Endocrinology & Metabolism 85:2822-7 2000.