Sequential Blood Collection from Inferior Vena Cava Followed by Portal Vein to Evaluate Gut Microbial Metabolites in Mice

Short-chain fatty acids (SCFA) are a major class of metabolites produced by the gut microbiota. Their critical roles in the interaction between the gut microbiome and other distant organs¹ have been supported by research describing how they modulate inflammation, signal through dedicated receptors, and serve as substrates in cellular metabolism^2,3,4,5. Recent work from our group has proposed that SCFAs are key in vivo inflammatory regulators of lung immune tone in vivo via the gut-lung axis^6,7. Additional reports have described their functional influence on metabolism via the gut-brain axis^8,9. Overall, the influence of SCFAs on host physiology and pathology is under active and intense investigation by numerous research groups spanning a wide range of disease processes.

Acetate (C2), propionate (C3), and butyrate (C4) are the primary SCFAs and are generated by gut microbiota through the fermentation of ingestible dietary fiber in the cecum and large intestine. All three SCFAs can also be obtained directly from the diet, and only acetate may also be produced by mammalian cells. SCFAs are absorbed in the colon and are partially utilized by intestinal epithelial cells (as an energy source, for local tissue immune modulation and to support gut barrier maintenance). They are also transported into the portal vein via the mesenteric venous system¹⁰. Butyrate is mainly consumed by intestinal epithelia, propionate by the liver^11,12, and acetate has been reported to act on muscle and adipose tissues after entering the peripheral circulation^13,14.

A comprehensive assessment of SCFA production, absorption, and functional activity requires knowledge of SCFA levels within the colonic lumen, in the portal circulation and peripheral blood. This can be accomplished by blood collection from portal vein (PV) and systemic circulation simultaneously or sequentially in the same animal. Since SCFAs are volatile¹⁵, measuring their levels in expelled fecal pellets may not accurately reflect levels within the colon. Furthermore, compared to measurements from colonic contents, the level of SCFAs present in the PV may more accurately reflect the net sum of the steady-state levels absorbed by the host versus the uneven levels produced by the gut microbiome throughout the length of the colon¹¹. These PV SCFA levels may thus be more relevant and appropriate for studying the effects of SCFAs on host physiology and pathology beyond the local effects within the intestine.

To perform the coordinated and near simultaneous collection of PV and systemic circulating blood, the diaphragm should remain intact so as to maintain normal blood circulation and support spontaneous breathing. Therefore, the inferior vena cava (IVC) presents an ideal site to obtain systemic circulation blood while collecting PV blood. This IVC blood can also be used for other purposes, such as measuring circulating cytokines to evaluate systemic inflammation.

Currently, only a few methods for collecting blood from both the systemic circulation and PV have been reported in larger rodents^16,17. Conventional methods, which require cannulation of vessels in rats, are technically difficult to perform in mice. In addition, the maximum amount of blood collected by these methods is usually no more than 0.3 mL¹⁸.

In this paper, we present a novel method that simplifies the process of dual blood collection from mouse IVC followed by PV in the same animal. The unique feature of the method is the ligation of the PV near the hepatic hilum just prior to PV blood sampling. This approach can expand the dimensions of the PV, thereby significantly improving the success rate as well as increasing the maximum collectable blood volume up to 0.5 mL.

All steps in this procedure were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of California San Francisco. The mouse gender and strain used were male C57BL/6J mice (weighing 25-35 g and aged 12-15 weeks). Female and/or other standard mouse strains can also be used.

1. Anesthesia

1. Wipe the bottom half of the mouse abdomen with 70% ethanol swabs. Administer anesthesia by giving an intraperitoneal injection of tribromoethanol (250-400 mg/kg). Ketamine/xylazine or isoflurane anesthesia can also be used. Apply ophthalmic ointment to the eyes to prevent corneal damage.
2. Inject buprenorphine (0.05-0.1 mg/kg) intraperitoneally for analgesia.
   NOTE: Buprenorphine is required because of the pain associated with the laparotomy procedure. 

2. Laparotomy

1. Place the anesthetized mouse on the surgical surface (with a heating pad underneath) in the supine position. Fasten all mouse limbs on the surgical board with tape. Confirm that the appropriate depth of anesthesia has been achieved by toe pinch without pedal withdrawal reflex.
2. Lift the skin and make a longitudinal incision with dissecting scissors along the midline from the umbilicus to the xiphoid process and caudally towards the base of the abdomen (external genital organ).
3. Make a similar incision into the peritoneum with dissecting scissors without perforating or damaging the intestines, diaphragm, or liver.
4. Expand the surgical field by making a full-length transverse incision (extending the width of the abdomen) with scissors on both sides of the umbilicus.
5. Move the large and small intestines to the surgeon's right side using a clean gauze soaked in warm PBS to expose the IVC below.

3. Ligature preparation for PV

1. Gently manipulate the intestinal segments within the surgical field with a cotton swab to locate the PV at the root of the mesentery.
2. Using a dissecting microscope to visualize the vessels and other abdominal structures, carefully pass a 7-0 prolene suture behind the PV near the liver hilum (between the duodenal vein and splenic vein)¹⁹ so that it is ready for ligation in step 5 (Figure 1).
   NOTE: Using a microscope from this step onwards is not mandatory but significantly improves the overall procedure's success rate.

4. Blood sample collection from the IVC

1. Insert the 28G ½-inch needle of a heparinized 1 cc insulin syringe into the IVC. Collect 0.2-0.3 mL of blood sample.
   NOTE: If too much blood is removed in this step, it will result in lower volumes of PV blood collected in step 5.
2. Leave the needle inside the IVC so as to avoid the bleeding that would occur with removal.

5. Blood sample collection from the PV

1. Gently tie the 7-0 prolene ligature around the PV; the PV should expand and dilate as it backfills. Insert the 28G ½-inch needle of another heparinized 1cc insulin syringe into the PV.
2. Aspirate slowly to collect up to 0.3-0.5 mL of blood from the PV. The PV may collapse but it should gradually re-expand with blood refilling from the mesenteric venous system.

6. Sample storage

1. Remove the needles from the IVC and PV after blood collection.
2. Transfer the blood samples from syringes to 1.5 mL heparinized eppendorf tubes and place them on ice for later processing into plasma for storage and analysis as needed.

Using the method described above, we can collect blood samples from the IVC and PV sequentially in the same mouse with a success rate of more than 95%. The average volumes of blood samples collected are 0.25 mL for the IVC and 0.35 mL for the PV.

Using gas chromatography-mass spectrometry (GC-MS), we measured the concentration of SCFAs in feces, PV blood, and IVC blood and were thus able to trace the absorption and transit of acetate (C2), propionate (C3), and butyrate (C4) from the colonic compartment where they are produced by fermenting gut bacteria, into the systemic circulation.

Specifically, in these experiments, we investigated the effects of dietary fiber intake on the production of SCFAs in mice. We provided mice with a high-fiber diet (35% fiber) for 1 week or 2 weeks and compared them with those fed a fiber-free diet. We expected and observed an increase in fiber-fermenting bacterial taxa in the high-fiber groups (data not shown). We next examined the colonic contents, PV blood, and systemic circulating (IVC) blood for levels of SCFAs (quantified by GC-MS at the Metabolomics Core Facility at the University of Michigan)³ (Figure 2). While we were able to detect a consistent increase in SCFAs with dietary fiber exposure in the PV samples, the SCFA measurements in the colonic fecal samples were far more variable from mouse to mouse compared to the PV and IVC blood measurements (colonic fecal matter versus PV/IVC blood for acetate, propionate, and butyrate in Figure 2). This variability may be due to the non-homogenous distribution of SCFAs in the fecal matter or may reflect variable absorption of SCFAs by the intestinal epithelium. Regardless, for the purposes of investigating the effects of SCFAs on host extra-intestinal physiology, levels within the PV and in the peripheral circulation are likely to be more relevant to study.

By comparing SCFA levels within the PV with levels in the systemic IVC blood, we were also able to observe levels of the absorption of acetate and propionate by the liver (Figure 2A,B). The proportion of acetate remaining in the systemic blood (compared to the PV) after passing through the liver is higher than that of propionate (more than 50% of PV acetate made it into the peripheral circulation compared to less than 30% for propionate). This is consistent with previous reports that propionate is absorbed or used by the liver, with lower levels entering the systemic circulation, whereas acetate may be absorbed or used in other peripheral tissues, such as muscle tissue^13,20 (Figure 2A,B). Nevertheless, previous work from several groups has shown that the levels of propionate that gets absorbed into systemic circulation is able to significantly influence physiology and pathology in distant organs^21,22,23,24.

In addition, we found that butyrate concentrations in both PV and systemic blood were much lower than propionate, although fecal concentrations of butyrate and propionate were quite similar. These findings confirmed previous work that gut-generated butyrate in the colon is largely absorbed and used locally by intestinal epithelia.¹¹ (Figure 2C).

Normally, during abdominal dissection, one observes that the dimensions of the PV are smaller than the IVC (Figure 3A). However, after the suture around the PV is tied (at approximately 4mm inferior to the hepatic hilum) and the vessel is ligated, the PV increases in size and is almost equal in diameter to the original size of the IVC before PV ligation (Figure 3B). As can be seen in the figure, the IVC slightly collapses after PV ligation (Figure 3B). Overall, this approach facilitates maximal collection of blood from the PV in mice. In our experience, collecting the PV blood first and then the IVC blood produced less optimal than results. This is because we observed significant collapse of the IVC after PV blood collection.

Figure 1: PV ligation. A suture is passed behind PV between duodenal vein and splenic vein. Please click here to view a larger version of this figure.

Figure 2: Concentration of short-chain fatty acids (SCFA) in feces, portal vein (PV) blood and systemic venous blood. Levels of SCFAs were measured for three different experimental mice groups including fiber-free diet for 2 weeks (n=10 mice), high-fiber for 1 week (n=9 mice) and high-fiber for 2 weeks (n=7 mice). (A) Acetate (C2), (B) Propionate (C3) and (C) Butyrate (C4). Levels of SCFAs in feces are reported as nM/mg (left Y-axis) of fecal material and in PV and inferior vena cava (IVC) plasma as µM (right Y-axis). Data in the figures are expressed as mean ± SD. Statistical analyses were performed using 2-tailed nonparametric Mann-Whitney analyses. p values < 0.05 were considered significant. p values are represented as follows: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. Please click here to view a larger version of this figure.

Figure 3: Mouse PV dilates and expands after tying the ligature at 4 mm below from the hepatic hilum. (A) PV and IVC before PV ligation; (B) PV and IVC after PV ligation. Please click here to view a larger version of this figure.

This paper describes an innovative in vivo method for near simultaneous collection of blood samples sequentially  from the IVC and PV in the same experimental mouse. This method is useful for measuring the levels of gut microbiota-generated products, such as SCFAs, that transit through the portal circulation. The average maximal volume of blood that can be collected during a terminal procedure in mice (weighing 25-30 g) is approximately 1 mL/mouse, which in turn represents 50% of the total circulating blood volume¹⁸. However, obtaining sufficient blood for desired experimental measurements from two interacting and neighboring venous systems can be challenging. The intravascular pressure of the PV in the mouse is only 3-5 mmHg higher than that of the IVC²⁵, so a decrease in pressure in the PV can lead to a decrease in pressure in the IVC and vice versa. In addition, the portal venous system in mice, which consists of multiple vessel branches from the mesentery and spleen²⁶, is relatively small and fragile. For these reasons, using a dissecting microscope for the accurate identification of the PV and its associated branches, and the prevention of PV and IVC collapse are critical to the success of this procedure.

The novelty of this protocol lies in the ligation of the PV prior to blood sampling. Blocking the PV drainage into the liver increases the pressure within the PV and therefore dilates it²⁷, which in turn facilitates blood collection as well as increases the amount of blood available for collection in the portal venous system. However, the ligation of the PV reduces the amount of venous return and then causes the IVC to collapse rapidly, making it very difficult to insert a needle into the IVC and limiting the amount of blood that can be collected. Therefore, our protocol recommends first aspirating blood from the IVC, before ligating and occluding the distal PV. Care must be taken to avoid collecting too much blood from the IVC as this can lead to significant and unrecoverable PV collapse even after PV ligation. In addition, ligating the PV also helps to avoid mixing PV blood with hepatic vein blood. Notably, the volume of blood samples collected in the protocol is sufficient for further studies such as metabolomic analyses, cytokine assays, and measurements of gut leakiness: 0.2-0.3 mL for IVC and 0.3-0.5 mL for PV.

The PV/IVC blood collection method presented here has some limitations. First, the blood collected from the IVC might not perfectly represent systemic circulating blood. It could overestimate the contribution from the lower part (caudal half) of the body as well as underestimate the contribution from the upper (cephalad) part of the body. Second, ligation of the PV may cause venous congestion in the bowel and other abdominal organs. Therefore, blood should be collected immediately after PV ligation to avoid collecting unwanted byproducts resulting from outflow obstruction from the mesenteric system and any incidental damage to the intestinal barrier. Third, it is not possible to collect blood samples serially at various time points to monitor biological temporal changes since this is a non-survival/terminal procedure. In contrast to cardiac puncture, a well-known and reliable method for obtaining systemic circulating blood, this method preserves an intact diaphragm with continued spontaneous breathing. In our experience, maintaining a diaphragm intact is critical for this method of sequential venous blood collection from the PV and IVC.