Location, Location, Location: Appraising the Pleiotropic Function of HMGB1 in Fetal Brain

We read with interest the paper by Zhang et al entitled “HMGB1 translocation after ischemia in the ovine fetal brain” (1). High-mobility group box 1 protein (HMGB1) is a fascinating protein with “many faces,” yet we are only beginning to uncover its roles in the developing brain (2). The findings of Zhang et al add to the much-needed body of literature in the field and validate our work in the same animal model and at the same near-term gestational age. Under intermittent umbilical cord occlusions (UCOs) insult with 3–4 hours of worsening acidemia and pH <7.00, we found an intracellular translocation of HMGB1 in neurons (3, 4), astrocytes (4), and microglia (5, 6) measured 24 hours post-insult. Interestingly, under control conditions, we found the dominant localization of HMGB1 to cytosol in neurons and astrocytes but not in microglia. Acidemic insult due to UCOs caused subtle and brain region-specific reverse shifts in neuronal HMGB1 patterns of cortical gray matter (3).

Our findings seem to stand in contrast to those by Zhang et al. A number of methodological differences in quantification of HMGB1 translocation distinguish the present paper by Zhang et al from our approach and may explain the contrasting findings in the dominant neuronal HMGB1 localization.

A stronger scenario of 30 min cerebral ischemia was modeled by Zhang et al with immunohistochemical (IHC) analyses performed at 48 and 72 hours post-insult whereas we attempted to mimic the generic human labor process with multiple repetitive, intermittent 1-min lasting global ischemia episodes and IHC at 24 hours post-insult. It is possible that with the more severe brain ischemia apoptosis pathways have been triggered and the nuclear HMGB1 translocated to the cytosol en route to act as the extracellular cytokine for microglia (7).

With the DAB chromogen stain used, the delineation between nucleus and cytosol without a counterstain is less clear than with fluorescence or dual/triple stains. It is not clear how the authors determined where to draw these lines except by visually recognizing the translocation. Rather than contouring around specific brain regions, the authors drew around where they thought they saw positive translocation, and then measured the percentage size of that outline relative to the whole brain contour. Next, the authors deliberately selected only cells showing evidence of HMGB1 translocation. An additional selection bias may have been difficult to avoid as the authors chose the already larger pyramidal cortical gray matter cells as “translocated” because they have large nuclei, in which they could see the translocation better. We suggest that an additional systematic approach of all layers needs to be undertaken with quantification of cytosol and nuclear HMGB1 signals. This may reveal additional translocated cells being the granular cells, which may seem as non-translocated because of their tight cytosol and dark nuclei. Lastly, formalin immersion fixation only penetrates tissue at a rate of approximately 1.6 mm/hour and after penetration it still needs to create crosslinks to fix proteins. We and others perfuse and then immerse. The fixation is then coming from two directions at once (i.e. inside vessels and outside brain) and reaches the tissues faster to stabilize them. The areas the authors outlined as having heightened HMGB1 cytosolic signal seem close to a ventricle where formalin would be entering from, and their high signal may be reflective of good perfusion there vs. deep within the tissue.

In light of the strong focus on the inflammatory roles of HMGB1 in the extracellular space in the last decade of research it is worth remembering that HGMB1 is also essential for life per se and for tissue regeneration processes (8). Perhaps in line with the many faces of HMGB1 during development and in various cell types in health and disease, HMGB1 nucleo-cytosolic distribution patterns are remarkably tissue- and time-specific (8–10). A predominantly cytosolic location of HMGB1 has been reported in developing brain, promoting neurite outgrowth and implicated in early neuronal development in vitro and in vivo (8,9,11). Thus, alterations in nucleo-cytosolic predominance of HMGB1 in developing brain may not only result in pro-inflammatory and neurodestructive effects due to its extracellular functions but also in disruption of physiological neurodevelopmental programs due to its intracellular, notably also cytosolic, functions, which are yet to be well defined in mammals.