What is Diabetes?

Sixteen Post Bumps for 2016 ...  No. 2

Update Notes & Summary


This is but one of the papers that turned up as I was delving deeper into the progression of Type 2 diabetes and what it really involves.  It was this work that began the research and blogging journey into sorting out what insulin resistance is, if indeed it does exist as it is commonly "understood" from the standpoint of peripheral tissue glucose metabolism.   In the words of the authors of the paper:
... for three decades, the beta cell dysfunction has been shaded by the theoretical construction of peripheral insulin resistance.
These researchers have access to roughly 170,000 records of everyone diagnosed with diabetes since 1942 in the jurisdiction of the Bucharest Registry of Diabetes.  This post discusses the common defect found in all forms of diabetes:  β-cells dysfunction.  A breakdown of the insulin assembly line resulting in reduced secretion of biologically active fully formed insulin, and increased secretion of inactive proinsulin precursor.

Originally Posted 2/17/2013

"the proinsulin disorder in diabetes is not 
only ubiquitous but it is also precocious"

As a second post setting up for future discussions on insulin resistance and it's role in diabetes (especially type 2)  I wanted to share the following paper with you here:

It is a long paper, almost 18 full pages of text, and more than 9 to list the 298 (yikes!) references. I will by no means attempt to address it in its entirety here, rather I shall focus on the section entitled "INCREASED PROINSULIN AS THE MAIN BETA CELL SECRETORY DEFECT".  But first, the abstract of this 2007 paper:
Based on our clinical and epidemiological data, we have sustained for a long time the unitary character of the various phenotypes of the diabetic syndrome. In this paper, we add several arguments sustaining that the unitary character of diabetes is related to a common primary defect in the function of the beta cell endoplasmic reticulum, leading to an inadequate processing of the two main secretory molecules: pre-proinsulin and pre-proamylin. The post-translational changes of these molecules might explain the main proapoptotic and anti-regenerative pathogenic mechanisms leading to a progressive decrease in the β cell mass/function. In our view, the increased proinsulin levels encountered in various diabetes phenotypes could be not only a marker of beta cell dysfunction but also could indicate the main β cell defect, suggesting also its location.
What I find compelling about Tirgoviste's group is that their hypothesis has evolved  through studying well over one hundred thousand cases of diabetes in humans:
Our hypothesis is based on 4 decades of clinical activity, having at our disposal the epidemiological data provided by the Bucharest Registry of Diabetes, which from 1942 till present included almost 170 000 patients of all ages, representing all the cases of diabetes identified in the population from a geographical area. Thus, we could identify several epidemiological, clinical, genetic and biochemical details which, put altogether, led us to a rather new interpretation of the pathogenesis of the various diabetes phenotypes, in the center of which is comfortably placed the pancreatic beta cell.

It was because of this paper that I wrote How a β-cell Works.   In short, at the ribosomes, the genetic code is "translated" into a protein called preproinsulin.  This is transported into the endoplasmic reticulum (ER) with the help of a signalling sequence that is cleaved in the process.  The remaining protein is now called proinsulin.  Proinsulin is "assembled" into insulin in the ER as it is folded, cross-linked, and one of the three major chains is cleaved.  The result is insulin and c-peptide.  What is shown at right is the proinsulin just before the c-peptide is cleaved.  These "processed" proteins are transferred to the Golgi apparatus where they are packaged into vesicles or granules.  In a controlled fashion, these vesicles migrate to the cell membrane and fuse with it to release their contents in a process called exocytosis.

The "money" part of this article is as follows:

  • Circulating proinsulin (and proinsulin:insulin ratio) is elevated in all forms of diabetes.
  • Proinsulin is elevated in first degree relatives and descendants of both Type 1 and Type 2 diabetics.
"the proinsulin disorder in diabetes is not 
only ubiquitous but it is also precocious"

The site of this defect in β-cell function is the ER.  This is described in some detail in the paper but I don't think we really need to get into that nitty gritty.   Whether some are inherently susceptible to β-cell-ER dysfunction or have some pre-existing dysfunction is not definitively known, but the prevalence of T2 associated with obesity argues strongly for the nutrient overload hypothesis put forth.  Those that may have some predisposition or deficiencies may well just "overload" their β-cell-ER sooner.  
... our hypothesis, is related not to a major alteration of a structural or functional molecule of the ER, but rather to a “degree of inflexibility” of the ER adaptability in the presence of an increased insulin demand, as for example in the presence of obesity.  A structurally normal ER, but under-dimensioned for an increased or even normal secretory traffic, will have a lower capacity of pro-molecules processing.  In the presence of a high molecular traffic through its secretory labirintic structure, this will allow the release towards the GA (the main site of secretory vesicles assemblage) of an increased percentage of un-split (or just partially split) proinsulin.
The incomplete "maturation"of insulin in a compromised  β-cell-ER leads to the formation of "immature secretory vesicles".  While these are apparently usually only present in the center of the cell, when these are produced in greater number these mingle with normal "mature" vesicles and are sent off for exocytosis/secretion which does not normally occur in properly functioning  β-cells.   Direct quote, bullet-pointed for clarity:  

The exocytose of this immature vesicles will have two consequences:
  • the first consequence is that the plasma levels of proinsulin will increase proportionally with the concentration of proinsulin inside the secretory vesicles while the proinsulin-to-insulin ratio (PI/I), (usually < 0.1) will increase, sometimes becoming supra-unitary [ > 1 ]
  • the second consequence is that immature vesicles will improperly react to the movement commands,such that the physiological insulin-secretion pattern (oscillatory secretion, biphasic response after a stimulus) will be attenuated or totally disturbed. 
We don’t know yet exactly the chronology of these disturbances (increased proinsulin vs disappearance of the first phase insulin secretion) but we know that they are both very precocious in the natural history of diabetes. 
This section and their hypothesis is summed up as follows:
  1. increased proinsulin 
  2. defect of proinsulin processing inside the ER
  3. immature secretory vesicles
  4. increased intra beta cell proinsulin as a trigger for beta cell autoimmunity
  5. the relationship between amylin / toxic amylin oligomers / amyloid deposits / beta cell apoptosis
[We recognize] there are still some unanswered questions regarding the chain of events that leads from the beta cell proinsulin defect in the ER to the decrease of the beta cell mass.

I would note points 4&5 are the events leading from "impaired glucose tolerance" and perhaps early stage frank diabetes to what we might refer to as the degenerative phase of diabetes.  In T1, the massive  and rapid autoimmune response  leads to an irreversible destruction of β-cells that cannot be compensated for.  Warning, imperfect analogy alert!!   β-cells are not like female eggs of which we are born with a fixed number of and lose with age (though this has been challenged recently).  They are more like male sperm that are generated from pre-sperm (in non-bio-techno-speak) pretty much throughout a man's life though often with diminishing capacity with age.   There is some turnover in most of our cells.  Apoptosis is also referred to as "programmed cell death" -- a normal process by which aging cells that may malfunction are "killed off" and replaced by younger functional ones.  It is premature and rapid apoptosis that is pathologic.  Thus, in T2's the amount of β-cell destruction estimated and reported in many reviews varies.  Some reports suggest destruction is already significant once frank diabetes is diagnosed, while others report enlarged β-cell mass, at least in early stages but accompanied by glycemic values (and other metabolic markers) in diabetic ranges.

So (4) suggests that elevated proinsulin within the β-cell may trigger an autoimmune response and (5) discusses a possible role of amylin in the process of apoptosis.  In other words, 4&5 are the part of the process tying their proposed primary defect -- increased proinsulin and immature vessicle formation -- to the pathology leading ultimately to the disease known as diabetes.

What is diabetes?  It is β-cell dysfunction that is sufficient to prevent the combined β-cell mass from properly trafficking glucose, amino acids and fatty acids in circulation.  I hear so often that diabetes is a disease of glucose intolerance.  This is simplistic and metabolically ignorant, as hyperglycemia is but one of multiple metabolic mahems that ensue.

It is almost "fact" that the difference between T1 and T2 is that in T1 you have near complete absence of insulin while T2 is often associated with hyperinsulinemia.   This has led to the near universal acceptance of the concept of "insulin resistance" as the primary characteristic of T2 diabetes.  I've spent a lot of time discussing IR, and in both the scientific community and guru-land there is much debate and controversy over whether hyperinsulinemia causes IR or IR causes hyperinsulinemia, and whether obesity precedes either or both or is the result thereof.

Up coming, I'm going to open up a big ol' can of cognitive dissonance whoop ass.  There is a boat load of evidence that insulin resistance, as we understand and discuss it, is either secondary to all of this and perhaps does not even exist!  Stay tuned!!


rodeo said…
This might sound excessively nerdy but I'm really looking forward to the next installment in this series. Thanks for doing this.
Gabriella Kadar said…
Okay, so Evelyn, does this analogy fit? A person with the genes for celiac won't ever develop a problem if there is absolutely zero exposure to gluten. A person with the genes predisposing for diabetes won't develop it unless there is the appropriate environmental stimulus. So there's an underlying abnormality that blossoms into pathology given the right set of circumstances which does not de facto include obesity.
Steven Hamley said…
Interesting blog post. It really argues against both 'T2D just being more insulin resistance' and 'beta cell burnout'. I'm looking forward to the next one.

Here's a paper you might be interested in, which says:
LIRKO mice don't develop T2D because their beta cells compensate and that insulin is a factor in that compensation
CarbSane said…
I've seen that paper, it is quite interesting. I have to look again and see if there is any info on proinsulin in LIRKO. That mouse is still post-prandial hyperglycemic but less so at six months but fasting hypoglycemic. I just came across a paper where just "resting" beta cells overnight restored normal GSIS in T2's.

@Gabriella -- Since GSIS is normal, however, I don't think that's a fair analogy. You almost never have total absence of GSIS and even on a low carb diet you still have proteins to deal with (not to mention basal fatty acids). Whereas gluten can be totally eliminated and avoided.

@rodeo: Nerds are some of my favorite people!

Thanks all for reading and commenting on this series. I intend to do them with or without feedback, but it's always nice to hear!
Diana said…
I am interested to hear your proof that IR doesn't exist. Not promising to understand it...but interested to hear.

"Insulin resistance" yields 6.1 million results.
Kindke said…
CarbSane have you seen this paper?


CarbSane said…
Yes. I think I even blogged about it, but damned if I can find it. Actually it may not have been the paper but her Banting lecture on the same topic. My chrono posts widget isn't working :-( so I can't find it that way. Ahh, I just found the Banting lecture. http://diabetes.diabetesjournals.org/content/61/1/4.long
I've got a lot of other things to get to, hopefully I can find where I addressed her hypothesis previously.
Diana said…
Kindke - that article looks fascinating but it is behind a firewall. Damn. Meanwhile I found this on Corkey:


When someone like Corkey says that overeating isn't the cause of obesity, I don't dismiss it lightly. But still ELMM enabled me to lose weight so I've picked my team and I'm sticking with it.

The article does make me wonder whether all these different types of diabetes aren't completely different diseases lumped under one name. But what do I know?
Unknown said…
Diabeetus not Diabetes

Gabriella Kadar said…
What does GSIS stand for? Possibly the environmental stimulus for T2 has to do with % of macronutrients in the diet? Out of balance and most of the time excess?
Anonymous said…
That's a good article about Corkey. ELMM still might be the best tool in the tool box, but (from the article):

'While it’s extremely unlikely that a single food additive could be the villain that causes obesity and diabetes, Corkey hopes to identify a handful of problem additives that could be removed from the food supply. She also knows that in the end, the additives themselves might not be the key. Rather, her research might lead to a new understanding of a metabolic pathway that could be modified to treat obesity and diabetes or to a better understanding of how individual people respond to certain chemicals.'

At least she's not saying that 'what you've been taught all along about overeating is a lie'!

Kindke said…
I have the full text , email me if you want it

The thing that most interests me about Corkey is she seems to be pushing the idea that hyperinsulemia is primary to diabesity, which ofcourse fits in with my view ( also bias ).

"The article does make me wonder whether all these different types of diabetes aren't completely different diseases lumped under one name. But what do I know?"

Thats very possible I think, physiology is unfortunately complex. However I was/am under the impression the major culprit behind T2D was insulin's failure to properly control liver glucose production, i.e. rampant foxo1 activity.
Gabriella Kadar said…
Or could vitamin K2 deficiency have to do with it? A eucaloric diet rich in vitamin K1 and K2 sources would preclude an imbalance in macronutrients.
CarbSane said…
Glucose stimulated insulin secretion. There is always glucose, and can even be spikes outside of eating (e.g.rigorous exercise can cause glycogen breakdown and fatty acids to be released) so the analogy to celiac is a poor one. Making insulin is normal. Making a lot of insulin can be normal but some cannot make as much as others apparently. Then it is a vicious circle.

I'll post up an article later where just resting the beta cells overnight (somatostatin) can restore normal postprandial secretion. Pretty convincing stuff.
BigWhiskey said…
Evelyn, this simply rocks!!!
CarbSane said…
Diana wrote: "The article does make me wonder whether all these different types of diabetes aren't completely different diseases lumped under one name. But what do I know?"

I think this is the compelling thing about this paper -- looking at all different cases of diabetes and a crap load of studies, this elevated pro-insulin is near universal. The evidence for hyperinsulinemia being primary is slim to nil, and is hopelessly conflated between basal and postprandial hypersecretion.

That was one of the points of the hyperinsulinemia and insulin suppressive posts. In one study, only 28% of obese were pp hypersecreters ... so what of the 72%?

I just came across a study that showed basal insulin levels have risen on average... by 2-5% for the average. Is this sufficient to explain the obesity epidemic? I don't think so.

OTOH, that same data shows we're eating 300+ cal/day more on average. Corkey's paper (that I presume is the similar to the 2011 Banting lecture paper) seems to be ignoring the obvious and looking for other possible reasons. I'm not sure why. Perhaps one of her various agents has caused hyperinsulinemia -- that still leaves the hurdle of demonstrating HI causes obesity. I just don't see it.

The redox signaling angle she calls "novel" is not novel. There are a ton of papers out there regarding ROS as signalling molecules in beta cells and others.
OnePointFive said…
"The article does make me wonder whether all these different types of diabetes aren't completely different diseases lumped under one name. But what do I know?"

Doctors tend to classify diabetes pragmatically by treatment. If someone needs insulin at diagnosis, is young and is thin, then they are probably T1. Everyone who can manage their diabetes with oral medication for a while is T2 .
Many adult onset T1s get misdiagnosed as T2s (especially if the onset is relatively slow as is often the case)so do people with MODY Within the category we label T2 I certainly agree that there are probably many different types/causes.

The WHO committee that reclassified DM in 1998 (ie into T1 (a and b), T2, gestational and others) also thought this way. When defining T2 they said
"There are probably several different mechanisms which result in this form of diabetes, and it is likely that the number of people in this category will decrease in the future as identification of specific pathogenetic processes and genetic defects permits better differentiation and a more definitive classification with movement into "Other types".
The 'others' classification already has several sub sections.
CarbSane said…
I'm curious what you think of this proinsulin. I've found a whole lot more info on early insulin treatment for T2 that restores glycemic control in only a few days to a week mostly. And many stay in remission for a year or more. Some of these interventions (continuous infusion under hospitalization) sound pricey, but compared to years of treatment and complications?

I'm also a bit troubled by how many assays for insulin are cross reactive for proinsulin and have seen numbers around 40% for overestimates vs. when more specific assays are used. Could hyperinsulinemia actually just be hyperactivity but relatively normal insulin levels after all?
Jane said…
You know insulin makes cells take up Mg, which could explain how insulin treatment works. Mg --> ATP --> repair of damage.
Wouldn't it be awful if Kruse was right about Mg.

But he's wrong about iron. Corkey says iron makes beta cells produce more insulin. Presumably because iron generates ROS which open/shut the ion channels involved. Iron overload is implicated in diabetes, which I don't think Corkey mentions.

Kruse tells people on a PALEO DIET to take EXTRA IRON.
kb said…
Evelyn, I am exited about this post and the follow-up! Re the possible diabetic/ chemical connection, is there any way I can get access to that study? I am a borderline diabetic and also noticed a weird reaction to some food that will give me insomnia and a low level agitation. It usually happens when I eat in restaurants, eat processed food (which I have not yet pinpointed which ones) or drink wine, which has way more additives than most people realize. It sometimes happens for no apparent reason though.

Also, is it possible to set up the more scientific posts with, in addition to the geek version which I feel much of the scientific stuff is anyway, a summary that would relay better to one without the extensive science lingo/knowledge? I love this stuff but can't follow it sometimes. It would also be helpful to have an abbreviation glossary somewhere on the site. It is sometimes hard to understand if you don't know what every abbreviation stands for. For example ELMM - eat less move more?

I may not always understand everything, but continue to love your site!
Jane said…
I've read the paper twice now and I agree it's very important indeed. Here are three quotes which I think sum up the story. It looks like the primary defect is in proinsulin folding as you'd expect.

'We mention that the access of the
specific convertases – PC3 and PC2 – to the dibasic
amino-acids, where the split of the proinsulin and
proamylin molecules takes place, is highly
dependent on a correct folding and packaging of
these molecules. This is why the inclusion of these
“marginally wrongly folded and packaged insulin
molecules” in the secretory vesicles will lead
finally the high plasma proinsulin concentrations
encountered in the various phenotypes of diabetes.' - p119

'From the study of Loos et al.139 carried out on a
large number of normoglycemic individuals (1697
Europid men and women) results a strong
association between four TCF7L2 SNPs (including
rs7903146) and proinsulin level. Thus the authors
showed that T (minor) allele of this SNP was
strongly and positively associated with fasting
proinsulin and 32,33 split proinsulin relative to
total insulin levels, but not to insulin-to-glucose
ratio (IGR) at 30–min oral glucose tolerance test.
The authors made the supposition that this beta cell
defect could be related with both the major genes
involved in proinsulin processing (PC3, PC2)
because these contain TCF binding sites in their
promoters.' - p120

'One of the strongest arguments in favor of the
proinsulin beta cell defect as a primary cause of
diabetes is represented by the identification of
increased proinsulin levels in offspring of both
type 1 148, 58, 86, 88, 91 and type 2 270, 73, 92 diabetic subjects.
The proinsulin secretory defect appears before or
concomitantly with other beta cell defects recorded
during the pre-hyperglycemic stage of diabetes and
manifested by the loss of the physiological insulin
secretion oscillations 95, 101, 191, 233–235 or the loss of
the first phase insulin response 92, 96-98, 102, 236.' - p124
CarbSane said…
I've read it several times now! Right now I'm going down the trail of a Yudkin paper that looked at measured insulin vs. proinsulin levels. One (I can only get abstract of, printed it out but lost link at the moment) stated that insulin levels measured by highly specific assay were 38% of those measured by RIA, and thus "hyperinsulinemia" in NIDMM may not be HI at all! That's mind blowing. I don't find a lot newer than mid-90's on this, however. This would be a huge bungle in science if this weren't addressed and sorted out, but I have yet to come across anything more recent about it.
CarbSane said…
Hi kb! I finally got that paper: https://docs.google.com/file/d/0Bz4TDaehOqMKeUNJUTN6UXd1QXc/edit?usp=sharing

I haven't had a chance to read this one. There are similarities to the Banting Lecture however. I don't think there's much if any evidence to support the IR = compensatory to prevent hypoglycemia hypothesis underpinning this paper.
CarbSane said…
p.s. There is a link to the acronyms page on the right sidebar. It may need some updating,but I do really still want to migrate my blog at some point which takes precedent over any more tinkering.
Jane said…
Wow. Well when you think about it, how could sick beta cells be making MORE insulin? More incorrectly processed insulin, yes perhaps.
Nigel Kinbrum said…
It's impossible to c/p from a Google Doc :-/ Corkey wrote:-
"Insulin resistance purportedly stimulates increased insulin secretion, interpreted as the body’s attempt to overcome the resistance, however there is no satisfactory explanation for how insulin resistance might stimulate insulin secretion."
How about:- increased IR -> increased HGP/decreased glucose disposal -> hyperglycaemia -> increased insulin secretion?
CarbSane said…
Now that's ridiculous about the C&P. You can save the PDF and C&P from there. Sigh ... just to make life difficult I suppose.

The paper(s) I've been discussing and am about to discuss seem to be coalescing around the idea that beta--cell dysfunction is primary. Beta cells are supposed to secrete insulin (in different fashion) in response to glucose, amino acids and fatty acids. Excessive chronic stimulation -- that appears to be basal in nature -- causes secretion of incompletely/improperly formed insulin which is essentially a deficit. Let's say a certain level of A causes an insulin response of 100 (whatever, making up a number). If the insulin secreted is only 70 good/active insulin, the level of A does not go down as expected, thus more insulin secretion is signalled ... and the beat goes on. Essentially the same level of A requires insulin response of 130 to get the 100 equivalent.
Jane said…
Evelyn, I just found a paper from 2008 which suggests that people who have a genetic predisposition for diabetes (meaning, for a high proinsulin-insulin ratio) also have lower beta cell glucose sensitivity.
'Beta cell glucose sensitivity is decreased by 39% in non-diabetic individuals carrying multiple diabetes-risk alleles compared with those with no risk alleles'

I'm thinking, this could explain why some obese people feel better when they eat LC. Obese people's beta cells have a lot of insulin ready to be secreted, apparently, and if they're not very sensitive to glucose they would wait until blood glucose is quite high and then perhaps, suddenly release a lot. This would mean reactive hypoglycemia, which is very unpleasant.