β-Cell lipotoxicity in the pathogenesis of non-insulin-dependent diabetes mellitus of obese rats: Impairment in adipocyte-β-cell relationships

BUMP NOTICE: 

I'm bumping this post due to some recent discussions on social media regarding the reversibility of Type 2 Diabetes and the role of low carbohydrate diets in the growing "curing diabetes" paradigm.  Recently Dr. Roy Taylor (the investigator I most closely associate with the crash diet, though this is an extensive group) and colleagues, who implemented a "crash diet" to cure diabetes -- by reducing pancreatic and hepatic (liver) fat concurrently with rapid weight loss -- published the following in BMJ:  Beating type 2 diabetes into remission.    The criteria are non-diabetic markers for a period following reversal of the progression.  Here is where various low carbohydrate interventions "fail", as many are able to maintain normoglycemia only by avoiding carbohydrates.  The question remains if this is effectively the same as other reversals.  I contend it is not, as normal pancreatic function -- specifically GSIS -- has not been restored.  Anecdotally (because we really don't have truly long term studies on LCHF following weight loss phase), many LCHFers see rising fasting levels, and one outspoken advocate, Tim Noakes, is among the many who remain on metformin to manage glucose levels.  The burning question, that remains unstudied by the advocates of low carbohydrate diets, is whether such diets will slow, halt or reverse the disease itself.  Here is where rodent data is NOT promising, especially when diabetes is viewed across the full metabolic milieu rather than the isolated metric of blood glucose. 

---

---

ORIGINAL PUBLISH DATE:  July 21, 2011


β-Cell lipotoxicity in the pathogenesis of non-insulin-dependent diabetes mellitus of obese rats: Impairment in adipocyte-β-cell relationships

Yes, this is a rat study, but it seems applicable in light of the recent Diabetes "Crash Diet" Cure and the reductions in pancreatic fat seen in the study.  The study and the pancreatic fat issue were discussed here.  Summary:  11 diabetics followed a 600 cal/day diet for 8 weeks and regained normal insulin secretion and glycemic control.  The restored glucose induced insulin secretion (GSIS) was attributed to reduction in pancreatic fat content.

The investigators in the current study had this to say in the introduction (bolded emphasis is mine):

... β-Cell failure is accompanied in human and rodent NIDDM by complete loss of glucose-stimulated insulin secretion (GSIS) and, in all rodent models thus far studied, by a parallel reduction in β-cells displaying GLUT-2, the high-Km facilitative glucose transporter. 

Long-chain fatty acids, which may be central to the development of insulin resistance in NIDDM, can stimulate basal insulin secretion and inhibit GSIS in isolated islets. This suggests a scheme that could account for the β-cell abnormalities in pre-NIDDM and NIDDM and explain the relationship between insulin resistance and β-cell dysfunction.  In this model the increase in plasma free fatty acids (FFAs) associated with prediabetic obesity causes both the insulin resistance and the matching hyperinsulinemia of the prediabetic state; subsequently a further increase in FFAs causes the β-cell unresponsiveness hyperglycemia that characterizes overt diabetes.

The present study tests this hypothesis in a rodent model of obesity-associated NIDDM that most closely resembles the human disorder, the Zucker diabetic fatty (ZDF-drt) rat.  NIDDM begins in almost 100% of obese male ZDF rats (fa/fa) between 7 and 10 weeks of age, while virtually all obese female ZDF rats (fa/fa) remain nondiabetic. This permits early identification of prediabetic and nonprediabetic littermates. It is, therefore, possible to test the propositions that the hyperinsulinemia of the compensated prediabetic phase results, at least in part, from a moderate rise in plasma FFA levels and that the later loss of GSIS is caused by a further increase in FFAs to a critical concentration that blocks the β-cell response to glucose.

In parts of this study I'll discuss include three groups of rats:

1. obese male ZDF prediabetic and diabetic rats (fa/fa)
2. obese female ZDF nondiabetic rats (fa/fa)
3. nonobese male ZDF littermates (fa/+ or +/+)
4. male Wistar rats.

{Note:  This study also involved some in vitro experiments involving culturing islets in various media.  I will address these in a future blog post as time permits.  However feel free to comment on them if there's something there of interest to you.}

  

The researchers bred the ZDF rats that are genetically predisposed to obesity (by 4 weeks) and all males of this strain become diabetic by 8-10 weeks.  At 6 weeks the male ZDF rats are considered prediabetic.  The females of the strain develop hyperinsulinemia and hyperlipidemia but do not develop hyperglycemia.  All rats were fed the same chow.

To test their hypothesis:  

The validity of the [our] hypothesis ... requires that significant hyperlipacidemia be present in prediabetic rats prior to the appearance of hyperglycemia. We therefore obtained weekly measurements of free FFAs and TGs in the plasma of obese male ZDF prediabetic rats, using female homozygotes and male heterozygotes as obese and nonobese controls. Sampling began at 5 weeks of age-i.e., 3-4 weeks before the onset of overt diabetes-and continued until 14 weeks of age, at which time overt diabetes had been present for at least 4 weeks.

Here are the results:   (Please note quotes from the paper have certain statistical values and citations removed for ease of reading, as I find this "clutter" distracting.)

o = lean, non-diabetic male        ● = obese,diabetic male       ▲= obese, non-diabetic female

A: Only the obese male ZDB rats exhibited hyperglycemia.  

B: Plasma triglyceride levels were elevated in all but the lean rats.


C: Plasma NEFA/FFA are elevated only in the obese diabetic males.


D: Islet triglyceride content only elevated in obese male ZBD rats.

... the most striking finding was in the prediabetic males; at 7 weeks of age,  2 weeks before the appearance of hyperglycemia, their mean FFA level had risen to 1.2 ± 0.05 mM, significantly greater than in the obese and lean controls, and was 1.9 + 0.06 mM at 10 weeks of age, several days after the onset of hyperglycemia. Thus, islets of prediabetic rats were exposed to higher plasma FFA levels than their obese and lean littermates for an average of 2 weeks before the onset of hyperglycemia. 

There was a significant correlation between the morning blood glucose levels and the plasma levels of FFAs of Fig. 1 at the onset of overt NIDDM plasma FFAs exceeded 1.5 mM in every rat.

💥EDIT:  What the above boils down to is evidence that the rise in NEFA precedes the development of frank diabetes as determined by the "end line" symptom of hyperglycemia.💥

In addition to triglyceride content, they measured other functions of the islets themselves including both basal and glucose stimulated insulin production, and GLUT-2 transporter activity.  GLUT-2 is the "sensor" or mediator of GSIS in the β-cell.

We reasoned that if fatty acyl CoA in islets increases in proportion to the plasma levels of FFAs, in the presence of sufficient glycerol 3-phosphate for reesterification of the fatty acids, secondary TG formation might take place in the islets. A reduction in β-cell glycerol-3-phosphate shuttle activity, which has been reported in two other animal models of NIDDM, would increase glycerol 3-phosphate, a requirement for TG synthesis. We therefore measured islet TG content. In islets of prediabetic ZDF rats, [they found]

• TGs had increased 4-fold between the ages of 5 and 8.5 weeks.

• An additional, abrupt 2-fold increase occurred by 9 weeks of age, when the mean glucose level had reached 10.6 mM.  

• At 10 weeks islet TG content reached a plateau that was 10 times higher than in 5-week-old prediabetic ZDF rats and 10-week-old lean littermates

• [At 10 weeks islet TG content was] 4 times higher than in 10-week-old obese nondiabetic female littermates. 

There was a significant correlation between plasma FFA and islet TG levels consistent with reesterification of fatty acids in islets. Islet TG content was also correlated with blood glucose levels.

Next, the researchers further tested their hypothesis by pair feeding (e.g. assessing ad libitum intake of lean  controls and feeding the same diet/amount to the study group rats) the obesity/diabetes prone rats and lean littermates.  They raised the rats as usual until 6 weeks (already obese by ~4 weeks) when the male ZDB rats became prediabetic.  For the next 6 weeks they pair fed them - e.g. restricted calories - to see if it would prevent the β-cell impairment their ad libitum fed brethren exhibited.  The result?  

... the moderate reduction in plasma FFAs was associated with marked attenuation of the entire phenotype of NIDDM-i.e., the hyperglycemia, hypertriacylglycerolemia, accumulation of fat in islets, and loss of β-cell GLUT-2 glucose transporter. Basal hypernsulinemia was reduced by 60%, to below the levels of obese nondiabetic female ZDF controls, and GSIS was preserved at about half of the normal level in nondiabetic controls. The loss of GLUT-2 was prevented.

💥EDIT:  In a nutshell, they were able to prevent a significant amount of the inevitable development of diabetes with calorie restriction.💥

So, not a complete prevention of diabetes, but a marked retarding of the progression of the disease.   Some excerpts from the discussion (bold emphasis mine):

These results in a rodent model of obesity and insulin resistance provide evidence for a "lipotoxic" cause of some of the β-cell abnormalities observed before and after the onset of NIDDM. First, the onset of progressively increasing hyperlipacidemia, beginning ~2 weeks before the loss of GSIS and onset of hyperglycemia, provides ample time for the postulated changes in β-cells to occur; without exception the plasma FFA concentration exceeded 1.5 mM and islet TG content was greater than 0.7 μg per islet just before the onset of hyperglycemia, both values were significantly greater than those of either control group  (Lipid droplets were detected in sections of diabetic islets.) 

Second, a moderate reduction of the lipid abnormalities in obese prediabetic rats by means of caloric restriction reduced the β-cell abnormalities of NIDDM, although interpretation of this particular finding is complicated by the multiple consequences of dietary restriction. ...

... By invoking a relationship in which elevated FFA concentrations concomitantly induce insulin resistance in target tissues and insulin hypersecretion in β-cells, one can explain how insulin secretion manages to match the level of insulin resistance and prevent the development of hyperglycemia as obesity progresses. (It is not clear whether the very earliest stage of hyperinsulinemia is FFA-driven.)  However, whenever FFA levels in prediabetic rats exceeded 1.5 mM, their β-cells apparently became incapable of a further increase in secretory function to parallel the rising insulin resistance; first-phase GSIS is abolished at this stage of the disease and hyperglycemia appears. 

The results suggest that obesity-related NIDDM is characterized by two defects:

(i) a primary resistance in adipocytes to the antilipolytic effects of insulin in obese animals, which causes the hyperlipacidemia that, in turn, induces insulin resistance in muscle and insulin hypersecretion and

(ii) impairment in the β-cell response to glucose.

It remains to be shown that these findings in ZDF rats are relevant to insulin resistance and NIDDM in humans. 

That last caveat is, of course, important to keep in mind when extrapolating the results of genetic model rodent studies to humans.  However, much research has been done on humans looking at, for example, the role of NEFA/FFA in basal hyperinsulinemia, etc.

💥EDIT:  This loss of GSIS and concurrent lack of glucagon suppression has been shown to be characteristic in the progression of NIDDM in humans.  See:  Insulin Secretion in the Progression of Type 2 Diabetes ~ First/Early Phase, The Cause of Hyperglycemia in Type 2 Diabetes💥 

This study demonstrated that only the rats that became diabetic experienced this rise in plasma NEFA that preceded the hyperglycemia.  This is something I've blogged on for over a year now.  I would note that this current paper is relatively old, 1994.  From Bierman to McGarry to Boden to Frayn, the fingers of researchers who actually work with this stuff points to the NEFA as the critical primary condition in diabetes:  hyperNEFAimea --> hyperglycemia.  

This study also shows that there is obviously a genetic component to the disease.  Some apparently retain adipocyte functioning better and/or have more resilient β-cells than others.  In humans, obesity is almost always associated with elevated NEFA as larger adipocytes lose their insulin sensitivity.  This seems to have the following impacts on circulating NEFA levels:

1. Increased basal levels due to adipocyte-IR and greater NEFA release
2. Increased post-prandial NEFA release do to failure of insulin-mediated suppression in the IR-adipocyte.
3. Increased post-prandial NEFA do to decreased "trapping" of dietary fatty acids released from chylomicrons by the IR-adipocyte   (EDIT:  I've seen increasing evidence that this may not be a significant contributor if at all.)

The major underlying theory of the carbohydrate/insulin hypothesis is that insulin is trapping your fat in your cells and causing fat accumulation.  If this is to be believed, then reducing dietary carbohydrate would increase each contribution to circulating NEFA above.  In caloric deficit, the significant increase in NEFA seems to be benign as we're burning the fatty acids for energy.  In these early phases, there can be no doubt that the immediate improvements in postprandial hyperglycemia and need for insulin favor carbohydrate restriction.  But as is pointed out time and again, once maintenance is reached, the VLC diet (and even a moderately LC diet for some) is necessarily a HF diet -- both by percentage and absolute intake.  This can only serve to sustain the elevated NEFA levels that produced the diabetes in the first place.  And this is why even Dr. Davis  now acknowledges that his VLC diet has his diabetes in remission but not cured.  There are, however, still people hanging their health on the hook that w/o carbs, fatty acids won't accumulate in your organs because of that G3P theory that just won't go away.  Perhaps Mr. Gary Taubes could come out more forcefully on his blog with the mea culpa on that one.  Yes, G3P is required, but just as gluconeogenesis creates what glucose we need, so too, glyceroneogenesis creates what G3P we need.  And in the end, you want that, because it also seems somewhat protective -- for a while anyway -- for the NEFA/FFA to be re-esterified.  The triglyceride content per se is not a problem at certain levels or when it is tapped into frequently (e.g. trained endurance athletes).   Ectopic (non-adipose tissue) deposition of triglycerides may well be the body's last line of defense in the overfed state.

I believe that the diabesity epidemic has merely revealed that more of the human population is susceptible to developing diabetes than previously thought, and/or simply that more of the human population is reaching that critical level of obesity that sets the whole ball in motion.  Still, a large percentage of the obese never progress to the point of loss of glycemic control.  But if you've been diagnosed or are prediabetic, that recent "crash diet" study demonstrated that, in humans, there's something to this whole lipotoxicity thing after all, and that perhaps the dramatic calorie restriction has a more dramatic impact at "flushing" fat from the pancreas.  The appearance of lipid droplets in the tissue sections in the current study lead credence to the theory that local fat accumulation in the pancreas may amplify the NEFA delivery to the β-cell.  I don't see how any serious person can merely reject/ignore the tremendous body of research out there on this.

At maintenance, a LC/HF diet is going to increase NEFA, and more importantly mute the natural rise and fall throughout the day on a mixed diet.  Will this be detrimental to all?  Obviously not.  Could it be detrimental to some?  When statistically zero humans have ever consumed such a diet in the context of relative prosperity (e.g. cheap, fatty meat in abundance requiring almost no effort to procure), I think it's hard to make the case that this is an unimportant consideration in deciding that such an extreme diet won't be detrimental to those pre-disposed let alone optimal for all.  On the other hand, we have billions of humans eating carbohydrate based diets in good health.  The SAD is high in fat, carb and processed mixed foods.  It clearly leads to obesity which leads to diabetes in those pre-disposed.  To not progress from prediabetic -> diabetic, or to reverse early onset diabetes the solution is to lose weight.  Reverse the obesity to a certain point and get those NEFA levels down to levels the pancreas can handle.  But once you've done that, it would seem prudent to adopt a diet that will prevent NEFA from exceeding the threshold for β-cell demise.  

If your diet only keeps glucose at levels below a glucotoxic threshold, you're only treating the symptom.  If it also keeps NEFA levels below lipotoxic levels, now you are treating the disease itself.  

A last comment on NEFA and dietary fat.  Does dietary fat "cause" diabetes per se?  No.  Not if it is not contributing to the chronic excess energy intake that leads to obesity (or fat accumulation in excess of some individual threshold).  But once a person reaches a certain metabolic state, dietary fat DOES contribute to NEFA levels when the adipocytes fail to adequately trap the fatty acids.  This can do nothing but compound the problem when one is in a weight maintaining or gaining state.  This is a bit of a shift in my thought process that was heavily influenced by Keith Frayn's recent Fatty Acid Trafficking paper dealing with post-prandial contributions to NEFA.  A high fat, very low carb diet would only further mute the proper suppression of NEFA release in the post-prandial phase, and if you truly believe that fat deposition in fat cells is reduced by the low carb state, dietary fat can only serve to increase NEFA further.