Enteroendocrine Reprogramming by Altered Epithelial-Mesenchymal Crosstalk in Metabolic Surgery

Abstract

Metabolic surgery is an effective treatment option for type 2 diabetes. However, the therapeutic scope has been limited by unexpected inconsistent outcomes. This study aims to overcome these obstacles by determining fundamental mechanisms from a novel perspective by analyzing and comparing the surgical anatomy, clinical characteristics, and outcomes of metabolic surgery, including duodenal-jejunal bypass, Roux-en-Y gastric bypass, biliopancreatic diversion, one anastomosis gastric bypass, and their modified procedures, predominantly focusing on nonobese patients to mitigate confounding effects from overweighted type 2 diabetes. Regional epithelial cell growth and unique villus formation along the anterior-posterior axis of the small intestine depend on crosstalk between the epithelium and the underlying mesenchyme. Due to altered crosstalk between the epithelium and the opposite mesenchyme at the anastomotic site, the enteroendocrine lineage of the distal intestine is replaced by the proximal epithelium after the bypass procedure. Subsequent intestinal compensatory proliferation accelerates the expansion of the replaced epithelium, including enteroendocrine cells. The primary reasons for unsatisfactory results are incomplete duodenal exclusion and insufficient biliopancreatic limb length. We anticipate that this novel mechanism will have a significant impact on metabolic surgery outcomes and provide valuable insight into optimizing its effectiveness in type 2 diabetes.

INTRODUCTION

Metabolic surgery is highly effective treatment option for the patients with type 2 diabetes. Nevertheless, the therapeutic scope has been restricted due to the unexpectedly inconsistent outcomes. This study aims to overcome these obstacles by determining fundamental mechanisms from a novel perspective.

Two independent factors are associated with the pathogenesis of surgical type 2 diabetes. These factors include excessive adipose tissue in overweight individuals and an imbalanced release of gut hormones such as incretins. Improving altered gut hormone secretion entails identifying the major target hormones responsible for metabolic changes and proper rerouting of nutrients. However, the main challenge lies in elucidating the underlying mechanisms of altered hormonal release from the gastrointestinal tract.

A previous study examining the density and hormonal gene expression of small intestinal enteroendocrine cells revealed several changes in their distribution following a Roux-en-Y gastric bypass (RYGB) [1]. These findings suggest a critical link between the development of postoperative hyperglycemia and the altered distribution of enteroendocrine cells following surgery. Therefore, we investigated the mechanism that maintains the spatiotemporal regulation of regionalized epithelial identity, as well as the factors that contribute to alterations in the distribution of enteroendocrine cells and the expression of hormonal genes. We analyzed and compared various metabolic surgeries, such as duodenal-jejunal bypass (DJB), RYGB, biliopancreatic diversion (BPD), and one anastomosis gastric bypass (OAGB), as well as their modified procedures, based on their surgical anatomy, clinical characteristics, and outcomes. This analysis aimed to clarify specific surgical patterns that may lead to conflicting results. We exclusively included nonobese patients to mitigate compounding variables from overweight individuals.

PHYSIOLOGY OF INTESTINAL EPITHELIAL REGENERATION AND IDENTITY ALTERATION

Each part of the small intestine has its unique structure of epithelial tissue and type of cells, which allows the distinctive functions necessary for efficient nutritional assimilation [2]. The modeling of the intestinal epithelium and villus formation depends on the 2-way communication between the epithelium and underlying mesenchyme, which is mediated by signaling pathways [3,4]. When the functional epithelial area is lost through resections or bypasses, the remaining intestine goes through morphological and functional adaptive compensatory responses. These include changes in the distribution of enteroendocrine cells and transcriptional activities [1]. New developments in single-cell RNA sequencing have shown that following proximal small bowel resections, the epithelium of the distal small intestine undergoes regional reprogramming to acquire a proximal identity [4].

The epithelial-mesenchymal crosstalk impacts epithelial cell growth and regional villus formation along the anterior-posterior axis (regionalization) of the small intestine [5]. Therefore, intestinal anastomosis after a bypass or resection might promote alteration of epithelial identity by the crosstalk between opposite epithelium or mesenchyme exposed at the anastomotic site. In addition, the mechanism is partly based on postoperative proliferation, which takes time; noticeable discrepancies become evident a few months after surgery in humans [6]. Ileal transposition, a procedure consisting of transposing an isolated segment of the ileum to the jejunum, which results in an alteration of the normal distribution of endocrine cells along the gut without changes in the total length of the intestine, is associated with a net increase in insulin sensitivity. Interestingly, these results were observed in rats 45 days after the operation and were likely ascribed to the time to proliferation of reprogrammed epithelium [7]. However, the clinical implication of epithelial identity alteration remains unclear. Nevertheless, the region-specific function of enteroendocrine cells in different parts of the small intestine leads to significant metabolic changes. Additionally, rapid turnover rate of the intestinal epithelium and compensatory proliferation facilitate the propagation and spreading of the altered identity to the distal intestine. Such as the striking resemblance of the alimentary Roux limb of jejunum to duodenum in a pylorus-preserving DJB [8].

ALTERED IDENTITY OF THE ROUX LIMB AND COMMON CHANNEL

The surgical DJB procedure has 2 modifications: pylorus-preserving (duodenal-jejunal anastomosis) and conventional (gastro-jejunal anastomosis) types. A duodenal-jejunal anastomosis was developed to preserve pyloric sphincter function but revealed differences in glycemic control. In the pylorus-preserving DJB group, there was a tendency toward high blood glucose concentrations, which led to unsatisfactory outcomes [9,10,11,12,13,14]. In contrast, recurrent hyperglycemia was rarely detected in the gastro-jejunostomy DJB group, which yielded favorable outcomes [15,16,17,18,19].

These findings imply that the duodenal epithelium, attached to the pyloric sphincter could potentially cause recurrent hyperglycemia. However, the area of the exposed duodenum was considered too small to induce hyperglycemia.

A comparison of surgical DJBs and endoscopic DJB liners (DJBLs) could demonstrate a possible mechanism. Unlike pylorus-preserving DJB, conventional DJB and DJBL demonstrate minimal blood glucose elevation. Comparing pylorus-preserving DJB with conventional DJB and DJBL (recurrent vs. non-recurrent), individually, suggests that duodenal exposure and anastomosis procedure are the possible factors that predispose to recurrence, respectively. Nevertheless, they do not cause recurrence, exclusively. These findings suggest that hyperglycemia may be associated with coexistence of nutritional exposure of the duodenal epithelium and combined anastomosis procedure.

Studies on the impact of transpositions elsewhere in the intestine, such as the transposition of the ileal segment to the jejunal area, the jejunal segment to the ileal site, the duodenal segment to the ileal area, and vice versa, have been conducted to evaluate the epithelial changes in the intestine. The common features of the adaptations are as follows: when the ileal segment is transposed to the jejunal area, the sizes of the villi increase to a similar extent as the jejunal villi; when the jejunal segment is transposed to the ileal site, the villi shrink to a length likely to be similar to the ileal villi; and when the duodenal segment is anastomosed with the ileum, the villus of the ileum increases in size to the extent of the duodenum [20]. Based on these findings, it is more plausible that the proximal intestinal epithelium migrates to the opposing intestine.

Jejunojejunostomy, another surgical anastomosis used in the DJB procedure, is not exempted from identity changes. The distal bilio-pancreatic (BP) limb affects the epithelial identity of the common limb in accordance with the same mechanism as previously mentioned. Although the primary role of the long BP limb is to separate nutrients from the duodenum and upper jejunum, another role is to provide more distal jejunum for common channel anastomosis to mitigate the proliferation of proximal epithelium to the subsequent common channel which promotes recurrence. In a recent report on the importance of the BP limb in metabolic surgery, the beneficial effects of the surgery disappeared after BP limb excisions in rats with improved hyperglycemia that underwent DJBs with a long BP limb [21], suggesting the presence of additional functions of the long BP limb beyond keeping nutrients away from the duodenum and upper jejunum. The cancellation of beneficial effects is likely induced by the migration and proliferation of the preceding duodenal or upper jejunal epithelium to the common channel.

Usually, gastrostomy feeding to the bypassed segment cancels the beneficial effect of RYGB (Fig. 1A). But surprisingly, the glucose tolerance dose was not impaired with the duodenal-jejunal transit procedure following RYGB (Fig. 1B). The preservation of the metabolic effect despite nutrients passing through the duodenum and upper jejunum has been mysterious [22].

Fig. 1. Different outcomes of RYGB with duodenal-jejunal transit and feeding gastrostomy tube. The beneficial effect of RYGB is canceled after gastrostomy feeding through the bypassed segment (A). Surprisingly, the glucose tolerance dose was not impaired with the duodenal-jejunal transit procedure following RYGB (B).

RYGB = Roux-en-Y gastric bypass.

However, it would be acceptable with applying the hypothesis that the epithelial lineage spreads from the proximal intestine to the distal intestine across the anastomosis. This implies that most of the characteristics of the duodenal epithelium are lost when the jejunal epithelium replaces the duodenal epithelial lineage.

DIABETOGENIC ROLE OF GLUCOSE-DEPENDENT INSULINOTROPIC POLYPEPTIDE (GIP)

Recently, synthetic dual-acting peptides targeting GIP and glucagon-like peptide-1 (GLP-1) receptors have significantly affected type 2 diabetes patients, highlighting their potential therapeutic benefits [23]. However, clinical studies regarding GIP administration remain inconclusive. Unlike GLP-1, GIP is considered a hormone targeted by the foregut hypothesis, as it induces hyperglycemia instead of insulin secretion. Researchers continue to discover new evidence supporting the hypothesis that a sufficient length of the BP limb enhances the effects of surgery [24]. Furthermore, the upregulation of GLP-1 and the downregulation of GIP are typical features of the BPD procedure, which is most effective in improving type 2 diabetes. Notably, the surgical characteristics of the BPD procedure is the complete exclusion of the duodenum and the long BP limb [7].

The surgical outcome of modified BPDs compared to classic BPD, demonstrates the importance of complete duodenal exclusion from nutrients and sufficient biliopancreatic limb length. The results of pylorus-preserving BPDs with a duodenal switch were not comparable to those of classic BPDs, even though bypassed BP limb length was the same as a classic BPD [25]. The remaining duodenal epithelium at the anastomosis site may have been responsible for the unsatisfactory outcomes. In addition, the metabolic outcomes of revisional surgery following classic BPD, where the gastric pouch is reduced in size, the common-channel limb is lengthened, and total duodenal exclusion is maintained with an undiversified BP limb length, were comparable to a classic BPD [26].

Ultimately, lengthening the BP limb reduces the number of K-cells exposed to nutrients, resulting in decreased GIP secretion and improved outcomes. Therefore, we can deduce that the length of the BP limb directly correlates with the postoperative decrease in GIP secretion. However, we can only accept the comparison by excluding other factors that impact GIP alterations, such as incomplete exclusion of the duodenum from nutrients and weight fluctuations in overweight subjects. Table 1 [7,27,28,29,30] demonstrates the propensity for an inverse correlation between postoperative GIP changes and the BP limb length ratio. Surgical patients with minimal obesity were eligible, and the applied procedures were those that do not allow nutrient exposure to the duodenum and upper jejunum, such as RYGBs, classic BPDs, and OAGB.

Table 1. Demonstration of the propensity for an inverse correlation between postoperative GIP changes and the BP limb length ratio.

Operation	AUC GIP		GIP alterationa (%)	BP limb (cm)	Reference
	Before operation	After operation
BPD	3,297.0 pmol/L	1,874.0 pmol/L	56	≥250b	Guidone et al. [7]
RYGB	48.67 ng/L−1·min−1	51.56 ng/L−1·min−1	105	30	Laferrère et al. [27]
RYGB	50.96 pmol−1·L−1·min−1	52.66 pmol−1·L−1·min−1	103	40	Laferrère et al. [28]
RYGB	30.2 ng/dL·10 min	27.0 ng/dL·10 min	90	100	Fellici et al. [29]
OAGB	184.0 pg/mL·min−1	98.0 pg/mL·min−1	53	200	Kim et al. [30]

GIP = glucose-dependent insulinotropic polypeptide, BP limb = bilio-pancreatic limb, AUC = area under the curve, BPD = biliopancreatic diversion, RYGB = Roux-en-Y gastric bypass, OAGB = one anastomosis gastric bypass.

^aGIP Alteration (%) = $\frac{\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{O}\mathrm{p}\mathrm{ }\mathrm{A}\mathrm{U}\mathrm{C}}{\mathrm{P}\mathrm{r}\mathrm{e}\mathrm{O}\mathrm{p}\mathrm{ }\mathrm{A}\mathrm{U}\mathrm{C}}\times 100$ ; ^bOr above.

As a result, the length of the BP limb is the major determinant of the inconsistency of GIP values after bypass surgery; specifically, the length of the BP limb is inversely proportional to the GIP concentration. However, despite separating the high K-cell density area from nutrients, there is an increased GIP value of RYGB with a short BP limb [27,28]. Since the density of GIP-releasing K-cells gradually decreases from the duodenum to the terminal ileum [31], the high K-cell density at the end of the short BP limb may stimulate the proliferation of K-cells in the common channel. The presumed GIP decline inflection point is around 100 cm in length distal to the ligament of Treitz [32].

CONCLUSIONS

The core mechanisms of diabetic improvements include separating the entire duodenum and upper jejunum from the nutrient and modulating epithelial identity alteration, induced by crosstalk between the epithelium and opposite mesenchyme at the anastomosis site. The compensatory proliferation of residual intestines promotes the propagation of altered epithelial lineages to the distal area. The mechanisms uncovered in our study offer valuable insights into optimizing the effectiveness of metabolic surgery in type 2 diabetes.