1. Liver does not completely blanch/only one lobe blanches/odd blanching (vascular appearance)/other blanching anomalies.

This is almost always caused by air bubbles entering the portal vein. Check the entire length of tubing after the bubble trap for occult air bubbles, particularly the section of the tubing leading to the cannula, as well as the actual cannula itself. Ensure that a slow drip is present during actual insertion of the needle, as this will push out occult air pockets in the bevel of the needle.

Other causes include improper cannulation (i.e., partial cannulation of portal vein), cannulation of a neighboring duct (evidenced by inflation of the pancreas), clotted blood (caused by allowing the animal to die too soon before cannulation- you generally have less than 30s after death, unless you injected heparin beforehand), or improper buffer osmolality (is your HBSS really 1x?).

2. Liver immediately swells (excessively) upon increasing flow rate after initial cannulation, even after IVC has been cut.

Check the osmolality of your HBSS, and make sure you are actually perfusing HBSS, and not H20. Make sure the blood has not clotted already- is there any liquid coming out of the IVC? Make sure you did not introduce air into the system after cannulation- did you compromise the portal vein at any location aside from the insertion point? Was there an air bubble in an earlier section of the tubing you missed?

3. Liquid drains from the insertion point in the portal vein, in addition to coming from the IVC.

This is likely to be caused by a nick in the portal vein, due to movement of the needle tip within the vein. If it is a minor leak, and you still see the majority of the liquid draining from the IVC, you should be OK. If, however, you do not see the liver swelling at all when you clamp the IVC (i.e. nothing within 5 seconds), the leak may be too severe for perfusion to continue effectively.

4. An organ besides the liver swells, and upon perfusion of the digestion medium, appears to be digested as well.

This organ is almost certainly the spleen, as the portal vein is formed from the junction of the superior mesenteric vein and the splenic vein. This is common when the portal vein is cannulated at too low (distal from the liver) a location, or when excessive systemic pressure forces perfusate into the splenic vein (blood clotting/congealing). Cannulate at a location more proximal to the liver, as this will bypass the splenic vein, which in turn directs the majority of the flow to the liver.

It has been our experience that so long as there is sufficient flow going to the liver, digestion of the pancreas/spleen, while obviously not desirable, should not have any negative impact on the actual isolation procedure. In such instances, digestion time may have to be increased to accommodate the lower flow.

5. Liver swells on its own after a few minutes of digestion

This is normal provided you have perfused roughly 50% or more of your target volume. If, however, you notice that total yield is high, but viability low, this may be an indication that your collagenase concentration is too high and/or you are using too harsh a collagenase. Remember that 100 CDU/mL is a concentration we and others have found to be optimal with batches of collagenase we have tested; your own results may vary. We have not, however, ever experienced an instance where known working collagenase (which had been stored and handled properly) suddenly went bad. If you know your collagenase was working last week, or even last month, look to other sources of error.

6. The perfusion started out fine, but the needle slipped (happens for any number of reasons). Can I re-cannulate?

Assuming that the initial cannulation and blanching were ideal, our experience is that re-cannulation can be perfectly fine, assuming you have the skill and luck needed to cannulate an already compromised, collapsed vein. The recommended course of action is to lower the flow rate back to 1-2mL/min and cannulate again as quickly as possible. Each successive attempt lowers your chances of success, as the portal vein is quite fragile to begin with. Do not be surprised if your success rate for re-cannulation is several-fold lower compared to cannulation of a fresh portal vein.

7. The liver does not appear digested, even after perfusing all of the collagenase solution.

First, confirm that the liver indeed is not digested well, particularly if you do not have extensive experience in isolation of primary hepatocytes. Excise it and tear it apart in your plate of digestion medium- if you do not see the medium become cloudy after tearing and shaking, and if the liver comes apart in solid chunks, there is almost certainly a problem with digestion.

Digestion problems are generally related to improper perfusion, assuming total collagenase concentrations are not drastically lower than 100 CDU/mL. Blanching of the liver must occur almost instantly and completely upon perfusion with HBSS- you should see the liver start to turn pale when you first cannulate; once the IVC is cut and the flow rate turned up, the liver should suddenly and completely turn pale. If it does not, you will almost certainly fail to digest it, no matter how much you attempt to force the solution in (i.e. increasing the flow rate, clamping the IVC).

Also, make sure that the DMEM you chose has sufficient CaCl2 (the one we recommend has ~1.8mM).

Temperature is almost never a culprit here- a few degrees below 37C should have negligible impact on digestion efficacy.

In order of importance, as concerns digestion:

  1. Blanching/initial cannulation
  2. Sufficient washing with HBSS/presence of 0.5mM EGTA in HBSS
  3. Presence of sufficient CaCl2 in digestion medium
  4. Concentration of collagenase
  5. Length of perfusion
  6. Flow rate
  7. Temperature
8. Yield and/or viability problems

It is nearly impossible to efficiently troubleshoot both yield and viability issues concurrently. Yield issues should be dealt with first, and then viability.

Low yield is almost always related to blanching and/or collagenase concentration. Blanching is covered in the previous section, so we will focus here on collagenase. Regardless of viability, your primary goal should be high total yield. This generally means >20-25 million cells for a 20-25g mouse. If your yield is consistently ~25% less than that, consider raising your collagenase concentration by 30%. If your yield is consistently "<"10 million, double the concentration of collagenase. If you still have issues, try a few other batches of Worthington Type IV collagenase. As a troubleshooting step, try Type I collagenase as a positive control for digestion-- you should have very high yield, and very likely will have low viability. If even 100 CDU/mL of Worthington Type I collagenase does not solve your yield problems, there is something fundamentally wrong with your system (i.e. EGTA concentration of HBSS is too low or too high, CaCl2 concentration of digestion medium is too low, cannulation technique is poor, etc).

If increasing collagenase does not bring your yield up to >20-25 million, the problem likely resides in the actual perfusion. Your flow rate may be compromised by air trapped in the tubing, or the flow rate may simply be too low. Always refer to the appearance of the liver when digestion is complete- is it soggy and enlarged, with the consistency of a wet paper towel? Is it difficult to excise (intact) due to its fragility? Does it readily tear apart and cloud the medium? Do you have a readily visible pellet after spinning down?

Once yield issues have resolved, viability, if problematic, can be dealt with by tweaking collagenase type and/or concentration. Flow rate and digestion times may need to be tweaked for much larger ("<"35g) or much smaller ("<"20g) mice, but their contribution to total yield and viability is small compared to the quality of the cannulation and collagenase type.

For working with significantly smaller or larger animals, make initial adjustments to the flow rate first, and only change the digestion time if absolutely necessary. Collagenase concentration should remain static, assuming you have a known working stock and concentration. We have scaled this protocol to rats by increasing the flow rate from 10mL/min to 22mL/min, with only a slight increase in digestion time (~10-15%) and no change to collagenase concentration (held at 100 CDU/mL). Remember, when isolating primary cells, time = quality; it should take no longer than 15-20 minutes from the instant you cannulate until you have excised the liver.

9. After spinning down cells, the pellet is substantial, but total yield is low

Having a large pellet does not guarantee that you will have high yield and/or viability. On the other hand, if you have almost no pellet, it DOES guarantee that you have few cells.

If the digestion was complete, and the previously listed criteria met, this should never be an issue. However, it is possible that the cells were sufficiently damaged (during digestion, filtering, and/or washing) such that the majority of the pellet is composed of cell particulate. If your hemacytometer contains almost entirely cell particles/debris, this is likely the case. If so, treat the cells more gently, and/or cut back on your collagenase concentration by 30%.

10. Hepatocytes do not attach to plate---or, hepatocytes attach, but do not appear healthy

It is entirely possible to have a batch of cells that have been damaged during the isolation process, yet do not take up trypan blue. These are most often larger, vacuolated cells that do not reflect light well. Refer to previous sections pertaining to yield and viability.

As it is nearly impossible to perform all steps aseptically, there is always the remote possibility of contamination. If you find that your cells are not recovering as they should, keep a close watch for bacteria.

The type of plate matters for cell health, due to the surface area:volume relationship, which magnifies the "edge effect" as microwells become smaller. In our experience, it is much easier to have healthy, confluent, homogeneous monolayers in 12-well and larger plates, but much more difficult in 24-well and smaller plates. This is almost entirely due to the fact that hepatocytes tend to crowd in the middle of smaller wells, creating a ring of cells around the edge of the well, and a massive confluence of cells in the middle of the plate, which tends to be less than healthy due to excessive crowding. We therefore recommend that you learn to culture in 12-well or larger plates, and scale down only when you are confident in your technique.

We have had limited success with standard tissue culture-treated collagen-coated plates, whether we coat them ourselves or purchase ready-to-use plates. However, initial tests suggest that UNTREATED (non tissue culture-treated) polystyrene plates may be a viable option. While these plates are generally used when cell attachment is not desired, you will be coating the bottoms with collagen, so this is a non-issue; however, the lack of tissue culture treatment means that the plastic is hydrophobic. This reduces the magnitude of the meniscus, and cells therefore behave essentially as they would in 12-well (or larger) plates. The charge of the plates could also be a factor- in any event, we suggest Corning # 3738. We are in the process of confirming this hypothesis and testing even smaller-well culture dishes.

Finally, keep in mind that no matter how proficient your technique, there will be animal-to-animal variability, such that some batches of cells will look better than others.

11. Hepatocytes look normal, but there are bare patches , particularly in the center and/or edges

Most commonly, this is due to insufficient shaking immediately after plating, and/or too much or too little volume used, and/or too many cells plated.

Make sure you thoroughly shake the plate in a linear fashion after plating.

Double-check the cell density by plating a single well, shaking it, allowing it to settle for 30 seconds, and looking at it under low (100x) power. You should strive for 60-70% confluence at this time, assuming >90% viability.

When coating plates with collagen, ensure that the entire bottom surface is exposed to collagen, and that any coated regions are not scraped by your pipette tip.

Finally, refer to question #10 for a discussion on cell size and plating issues.

12. There are many dead cells stuck to live hepatocytes
  1. Improve your technique so that viability >90%
  2. Wash cells within 45 minutes after initial plating, to prevent any dead cells from having enough time to stick.
  3. This is unavoidable to some extent, and generally not problematic.
13. Hepatocytes do not evenly attach in smaller (i.e. < 12-well) plates/cells tend to congregate in the middle of the plate and around the edges, forming a "bullseye" pattern.
  1. This problem potentially encompasses two issues. First, we have observed that scaling hepatocytes into smaller wells requires higher-quality cells. Second, the narrower diameter of smaller wells increases the capillary effect, creating a larger meniscus that tends to pull cells towards the center before they have a chance to settle evenly.
  2. The issue of cell quality is a complex one, and is addressed on other parts of this protocol. A solution for the latter issue, uneven cell distribution/bullseye effect, is to use NON tissue-culture treated plates (i.e. Corning #3738, Costar(R) 24 well not treated multiple well plates). Standard "Tissue culture treated" plates undergo a process that makes the plastic more hydrophilic, while NON tissue culture treated plates retain the intrinsic hydrophobic properties of the plastic. Since the bottom of the plate will be coated with collagen, the lack of tissue culture treatment has no bearing on cell attachment; it it the side walls we are concerned with. The increased hydrophobicity appears to greatly reduce the "bullseye" effect, and in our experience, allows for an even monolayer of hepatocytes. Note that the non-treated plates also have different charge properties, and we cannot say for sure whether this plays any role; in any event, if you find that standard tissue culture treated 24 (or 96)-well plates make even plating difficult, try the non tissue culture treated versions.
14. Hepatocyte phenotype rapidly degrades after the first day in culture.

This is a common issue in two-dimensional culture, and may represent an inherent limitation of this system. In our experience (and corroborated by colleagues in other institutions), hepatocytes de-differentiate into fibroblastic monolayers after 24-30 hours of plating. A common misconception is that hepatocytes spontaneously die, allowing contaminating fibroblasts to take over; primary observation clearly demonstrates that this is not the case.

There may be ways to prolong the useful life of hepatocytes in 2-D culture, including plating on thick layers of gelatinous matrices and/or addition of specific growth factors. We have not extensively tested these approaches, although very preliminary data indicate that Matrigel and similar products may be useful, particularly when cells are embedded/sandwiched. Regardless of culture length and the measures used to preserve phenotype, it is critical that function be validated at pre-determined culture lengths, as gene expression is known to change rather dramatically over the course of 12-24 hour intervals, as do functional parameters; we are unsure whether functional output changes primarily as a consequence of gene expression, or if gross shifts in higher-order processes secondary to the extracellular environment/matrix have a dominant effect.