Secondary Logo

Journal Logo


Cochlear Pericytes Are Capable of Reversibly Decreasing Capillary Diameter In Vivo After Tumor Necrosis Factor Exposure

Bertlich, Mattis; Ihler, Friedrich; Weiss, Bernhard G.; Freytag, Saskia†,‡; Strupp, Michael§; Canis, Martin

Author Information
doi: 10.1097/MAO.0000000000001523


Local blood flow is a critical parameter for cochlear function. Impairment of cochlear blood flow itself or its regulation has been associated with numerous pathologies of the inner ear, including, but not limited to, sudden sensorineural hearing loss (1–3), noise-induced hearing loss (4) or Ménière's disease (5–7).

To this point, it has commonly been assumed that capillary blood flow of most tissues is mainly regulated by the precapillary small arteries and arterioles (8), like the spiral modiolar artery in the cochlea (1). However, the role of capillary pericytes has recently been re-evaluated in numerous tissues. Pericytes are, generally speaking, cells that adhere to the outer walls of the capillaries (9) and fulfil a broad range of functions. They form physical barriers, like the blood–brain (10) or the blood–retina (11) barrier, play a role in tissue regeneration (12), and contribute to the stabilization of microvasculature (13,14). Moreover, pericytes are, at least partially, able to contract and thus decrease capillary diameter (9,14–17), eventually contributing to the short-term regulation of local blood flow (9,18–20).

Especially, the capillary pericytes of the brain have been subject to recent investigations: it has been found that neuronal capillary pericytes are not only among the first structures to actively increase local blood flow in times of increased oxygen demand by dilation (16), but also to be among the first cells to suffer from hypoxemia and thus contribute to a persistent decrease in local microcirculation in cerebral ischemia (17,19–21).

It has been accepted that decreases in cochlear blood flow are the common final pathophysiological pathway of numerous inner ear diseases (1–3,22–24). Many of these pathologies seem to rely on the tumor necrosis factor (TNF) pathway to mediate their effects, including sudden sensorineural hearing loss (1,25) and noise-induced hearing loss (4,25). Moreover, there have been numerous reports that antagonization of tumor necrosis factor by the application of etanercept, a fusion protein that is used to competitively bind the tumor necrosis factor receptors (26,27), is considered to be beneficial in some of these pathologies (1,4,25,28).

Hence, we decided to investigate the effect of tumor necrosis factor on cochlear pericytes, their ability to affect capillary diameter and the potential of etanercept in revoking aforementioned effects.


All of the experiments reported were approved according to local regulations by the responsible authorities (Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit LAVES, Oldenburg, Federal Republic of Germany) under the license no. 33.9-42502-04-14/1427.


Animals used were albino Dunkin–Hartley guinea pigs specifically bred for experimental use and were purchased from authorized retailers (Harlan Laboratories, Ober-Ramstadt, Hesse, Germany and Charles River Laboratories, Sulzfeld, Germany). The body weight ranged from 200 to 450 g. Anesthesia was induced by an intraperitoneal injection of 50 mg/kg bodyweight (b.w.) ketamine and 5 mg/kg b.w. xylazine and sustained by repeated intramuscular injections of 25 mg/kg b.w. ketamine and 2.5 mg/kg b.w. xylazine every 30 minutes.

Surgical preparation lasted approximately 90 to 120 minutes. After the experiments were conducted, animals were euthanized by an overdose of anesthesia and subsequent cervical dislocation.

Surgical Approach

Intravital microscopy using fluorescein-labeled dextrane for the investigation of cochlear microcirculation was initially described in 1987 by Nuttal (29). He was also among the first to describe diaminofluorescein-2-diacetate as a selective marker for cochlear pericytes (15).

After the induction of anesthesia, a cervical venous catheter was surgically implemented. The external ear and bone covering the auditory bulla were carefully removed. By doing this, a free view on the lateral cochlear wall was achieved. After this, small periosteal vessels were removed using a microsponge. After the removal of the vessels, a small rectangular window of approximately 500 × 500 μm was carved into the cochlea at the second turn, exposing the stria vascularis. Afterward, fluorescein-labeled dextrane (molecular weight 500,000; 0.05–0.1 ml of a 5% solution in 0.9% NaCl; Sigma-Aldrich, Deisenhofen, Germany) was applied intravenously, allowing direct visualization of intravascular blood flow in the stria vascularis under illumination with a Leica EL6000 light source (Leica Microsystems, Wetzlar, Germany). After a clear view of the vessel window had been obtained, the bulla was filled with a solution of 5 mM 4,5-diaminofluorescein diacetate in dimethyl sulfoxide diluted 1:10 with sterile saline solution. After 20 minutes, the bulla was repeatedly washed with sterile saline, and images were obtained. If at least six pericytes were clearly identified, treatment continued; otherwise, the experiment was aborted. Images were obtained with a Leica M205 FA stereomicroscope (Leica Microsystems, Wetzlar, Germany). The proprietary Leica Application Suite software was used to record and save images for later off-line analysis. Quantification of blood velocity and vessel diameter was done by Cap-Image (Dr. Zeintl Biomedical Engineering, Heidelberg, Germany), a software specifically designed for this purpose (30). A schematic of where capillary diameter was measured can be observed in Fig1C and A sample of the videos recorded is available with the online supplemental material (

FIG. 1
FIG. 1:
A, Schematic of the timeline of each individual experiment. B, Distribution of the guinea pigs into individual groups. Numbers on the right indicate the number of pericytes measured in each group. C, Schematic of how diameters were measured at pericyte sites (continuous line) and downstream controls (dotted line). ∗ indicates pericyte soma, the arrow indicated the direction of blood flow.

Treatment Protocol

Twelve animals were randomly assigned to one of three groups. A schematic of the general course of the experiment as well as the treatment of the individual groups can be observed in Figure 1A and B. After a clear image of the stria vascularis was obtained and a minimum of six pericytes were visible, basal values were recorded. After that, either tumor necrosis factor or a placebo were applied topically for 20 minutes. After the application, the bulla and stria vascularis were rinsed for approximately 10 minutes and images were obtained again.

Finally, either etanercept or a placebo were topically applied for 20 minutes again, the bulla was then rinsed with sterile saline for 10 minutes and afterward, final images were obtained. After the acquisition of the final images, the animals were euthanized.


At three timepoints (basal values, after first treatment, after second treatment), which were 30 minutes apart from each other, images of the stria vascularis with visualized pericytes were obtained. During off-line analysis, two parameters were quantified: the capillary diameter at each site of a pericyte soma (μm) and capillary diameter downstream of these pericyte somas (μm) as a control (see also Fig. 1C). The values reported are relative change in capillary diameter compared with baseline ± standard deviation. The absolute values for the capillary diameters can be found in the online supplemental material (

To detect significant differences, we fitted linear mixed models that included a random effect for the animal and were estimated using a restricted maximum likelihood approach. A p value < 0.05 was considered to be significant. The software used for this was Project R (Build 3.2.5 for Windows, The R Project for Statistical Computing,


Capillary Diameter at Sites of Somas of Pericytes and Downstream Controls After Initial Application of Tumor Necrosis Factor or Placebo

Overall, n = 199 pericytes measured in 12 animals were considered. Of these, 141 pericytes were treated with tumor necrosis factor and 58 were treated with placebo; this disparity is owed to the fact that eight animals in two groups were initially treated with TNF. Since the treatment was biologically the same, the group were pooled.

After the application of placebo, the vessel diameter at pericyte sites remained the same, only marginally increasing by 0.3 ± 2.0% while treatment with tumor necrosis factor resulted in a decrease in diameter at sites of pericytes of 3.6 ± 4.3%. At downstream control sites, application of placebo lead to a marginal increase of diameter of 0.4 ± 2.5% while application of tumor necrosis factor caused a diameter decrease of 2.3 ± 2.9%. The fitted linear models showed a significant difference between treatment (TNF) and placebo (p < 0.001). There was also a significant (p = 0.002) difference between sites of pericyte somas and downstream controls (Fig. 2).

FIG. 2
FIG. 2:
Relative diameter change compared with baseline at sites of pericyte somas and downstream controls after initial application of tumor necrosis factor or placebo. • = outliers ∗ = p < 0.05.

Fraction of Contractile Pericytes After Initial Application of Tumor Necrosis Factor

Of the 141 pericytes that were initially treated with tumor necrosis factor, 39 (27.7%) showed a decrease in diameter that was >8.0% from the initially recorded basal value.

Capillary Diameter at Sites of Somas of Pericytes and Downstream Controls After Application of Etanercept or Placebo

Overall, 199 pericytes were considered. Of these 199, 58 pericytes in four animals had been treated with placebo twice, and 141 pericytes in 8 animals had initially been treated with TNF. Of these 141 pericytes, 82 were treated with placebo after the application of TNF, while 59 were treated with etanercept after initial TNF treatment.

Twofold application of placebo caused a negligible increase of diameters of 0.0 ± 2.7% at pericyte soma sites as well as a marginal increase of 0.2 ± 2.7% at respective downstream control sites.

Application of placebo after the application of tumor necrosis factor caused a minimal increase of diameters of 0.4 ± 2.4% at pericyte soma sites and an insignificant decrease of 0.4 ± 2.5% at respective downstream control sites.

Finally, after the application of etanercept following the application of TNF, an increase in diameter of 3.3 ± 5.5% was observed at sites of pericyte somas. Downstream controls showed an increase of 1.8 ± 5.5%.

The fitted linear mixed models showed a significant difference between treatment (Etanercept) and placebo (p = 0.021). We also demonstrated a statistically significant difference effect between treatment with pericytes and downstream controls (p = 0.029, Fig. 3).

FIG. 3
FIG. 3:
Relative diameter change compared with baselines at sites of pericyte somas and downstream controls after the application of etanercept or placebo after the initial application of tumor necrosis factor. • = outliers = p < 0.05.


First, we have been able to show that cochlear pericytes are capable of decreasing capillary diameter at sites of pericyte somas upon a physiological stimulus. Mere contractions of cochlear pericytes have already been shown by Dai et al. in 2009 (15). In this work, the proportion of pericytes that were able to show an active reduction of capillary diameter ranged from approximately 20 to 40%. Fittingly, our results are in line with the results reported.

However, in the aforementioned work, contractions were only observed under nonphysiological conditions such as very high extracellular concentrations of electrolytes like potassium or calcium. In contrast, this is the first time that a significant reduction of capillary diameter by cochlear pericytes has been shown after a physiological stimulus. The fact that tumor necrosis factor seems to play a major role in many inner pathologies, including acoustic (4) or physical (28) trauma as well as sudden sensorineural hearing loss (1) and that most of these pathologies coincide with impairment of cochlear blood flow (1,4) suggests that impairment of cochlear blood flow is at least partially mediated by active contractions of cochlear pericytes.

Since it has been shown that neutralization of tumor necrosis factor by etanercept (1,4,25) and blocking of sphingosine-1-phosphate (S1P) signaling (1,31) are effective in preventing tumor necrosis factor-related decreases in cochlear blood flow, it seems likely that the vascular effects of tumor necrosis factor on pericytes are at least partially mediated by S1P signaling. Fittingly, TNF-S1P-signaling has already been described to cause vascular contractions (32). In addition to this, defects in the sphingosine-1-phosphate receptor 2 are associated with a rapid degeneration of the stria vascularis as well as severe hearing loss, suggesting an integral role of sphingose-1-phosphate-signaling for stria vascularis preservation (33,34).

Moreover, we have been able to show that cochlear pericytes are not only able to actively reduce capillary diameter, but to dilate again if the stimulus for contraction (tumor necrosis factor) is neutralized by topical application of etanercept. The fact that cochlear pericytes are not only able to contract, but to relax again, is new and—to the best of our knowledge—has not been shown in any scientific publication so far. This observation is in line with previous studies that have found tumor necrosis factor-induced decreases in cochlear microcirculation can be revoked upon neutralization of tumor necrosis factor by application of etanercept (31).

In addition, clinical effects of successful therapy with etanercept are often associated with improved microcirculation (35–37).

Overall, the facts presented in our work strongly suggest that cochlear pericytes play an active role in regulating local blood flow, much alike the pericytes in other tissues.

Taking this assumption even further, one might postulate that cochlear pericytes have a very similar function to cerebral pericytes as early mediators of cochlear blood flow. Not only do cochlear pericytes show what seems to be a bidirectional control of capillary diameter at sites of pericyte somata like the cerebral pericytes do (16). Cochlear pericytes could also play a key role in one of the most common inner ear pathologies, sudden sensorineural hearing loss (SSNHL): it has often been postulated that SSNHL is of vascular origin, since it is like the retinal vein thrombosis, a pathology of clearly vascular origin, usually one sided and has similar risk factors (38). Increased fibrinogen levels have been known to reduce cochlear microcirculation and increase hearing threshold levels like it is observed in SSNHL (2). Decreased cochlear blood flow is also known to coincide with decreased pO2 levels in the cochlea (3), suggesting a similar ischemic damage dealt to the cochlea and its microvasculature to that observed in the ischemic brain (17).

This view is further supported by the fact that we have shown etanercept to be capable of revoking previously induced contraction in cochlear pericytes. Since etanercept has also been suggested to be effective in the therapy of SSNHL (1), there is strong evidence that pericytes play a major role in SSNHL, possibly contributing to the continuous decrease in local microcirculation in a vicious cycle of ischemia and hypoemia.

In conclusion, we have been able to show that cochlear pericytes are capable of contraction, which has been reported previously. However, we have been able to show that this contraction takes places after the exposition to a physiological stimulus and that this contraction is clearly reversible. These new findings make it very likely that cochlear pericytes contribute to the regulation of cochlear blood flow, much alike the pericytes of the central nervous system, where they play an integral role in health and disease.


The authors thank Dr. Lutz Freytag, CFO of DB Schenker Logistics for valuable aid with the statistics.


1. Scherer EQ, Yang J, Canis M, et al. Tumor necrosis factor-alpha enhances microvascular tone and reduces blood flow in the cochlea via enhanced sphingosine-1-phosphate signaling. Stroke 2010; 41:2618–2624.
2. Ihler F, Strieth S, Pieri N, Gohring P, Canis M. Acute hyperfibrinogenemia impairs cochlear blood flow and hearing function in guinea pigs in vivo. Int J Audiol 2012; 51:210–215.
3. Lamm K, Arnold W. The effect of blood flow promoting drugs on cochlear blood flow, perilymphatic pO(2) and auditory function in the normal and noise-damaged hypoxic and ischemic guinea pig inner ear. Hear Res 2000; 141:199–219.
4. Arpornchayanon W, Canis M, Ihler F, Settevendemie C, Strieth S. TNF-alpha inhibition using etanercept prevents noise-induced hearing loss by improvement of cochlear blood flow in vivo. Int J Audiol 2013; 52:545–552.
5. Bertlich M, Ihler F, Sharaf K, Weiss BG, Strupp M, Canis M. Betahistine metabolites, Aminoethylpyridine, and Hydroxyethylpyridine increase cochlear blood flow in guinea pigs in vivo. Int J Audiol 2014; 53:753–759.
6. Bertlich M, Ihler F, Freytag S, Weiss BG, Strupp M, Canis M. Histaminergic H3-heteroreceptors as a potential mediator of betahistine-induced increase in cochlear blood flow. Audiol Neurotol 2015; 20:283–293.
7. Nakai Y, Masutani H, Moriguchi M, Matsunaga K, Kato A, Maeda H. Microvasculature of normal and hydropic labyrinth. Scanning Microsc 1992; 6:1094–1097.
8. Fernandez-Klett F, Offenhauser N, Dirnagl U, Priller J, Lindauer U. Pericytes in capillaries are contractile in vivo, but arterioles mediate functional hyperemia in the mouse brain. Proc Natl Acad Sci U S A 2010; 107:22290–22295.
9. Attwell D, Mishra A, Hall CN, O’Farrell FM, Dalkara T. What is a pericyte? J Cereb Blood Flow Metab 2016; 36:451–455.
10. Zlokovic BV. The blood-brain barrier in health and chronic neurodegenerative disorders. Neuron 2008; 57:178–201.
11. Pfister F, Przybyt E, Harmsen MC, Hammes H-P. Pericytes in the eye. Pflugers Arch 2013; 465:789–796.
12. Kramann R, Humphreys BD. Kidney pericytes: Roles in regeneration and fibrosis. Semin Nephrol 2014; 34:374–383.
13. von Tell D, Armulik A, Betsholtz C. Pericytes and vascular stability. Exp Cell Res 2006; 312:623–629.
14. Bichsel CA, Hall SRR, Schmid RA, Guenat OT, Geiser T. Primary human lung pericytes support and stabilize in vitro perfusable microvessels. Tissue Eng Part A 2015; 21:2166–2176.
15. Dai M, Nuttall A, Yang Y, Shi X. Visualization and contractile activity of cochlear pericytes in the capillaries of the spiral ligament. Hear Res 2009; 254:100–107.
16. Peppiatt CM, Howarth C, Mobbs P, Attwell D. Bidirectional control of CNS capillary diameter by pericytes. Nature 2006; 443:700–704.
17. Hall CN, Reynell C, Gesslein B, et al. Capillary pericytes regulate cerebral blood flow in health and disease. Nature 2014; 508:55–60.
18. Hamilton NB, Attwell D, Hall CN. Pericyte-mediated regulation of capillary diameter: A component of neurovascular coupling in health and disease. Front Neuroenergetics 2010; 2: pii:5.
19. Fernandez-Klett F, Priller J. Diverse functions of pericytes in cerebral blood flow regulation and ischemia. J Cereb Blood Flow Metab 2015; 35:883–887.
20. Dalkara T, Alarcon-Martinez L. Cerebral microvascular pericytes and neurogliovascular signaling in health and disease. Brain Res 2015; 1623:3–17.
21. Yemisci M, Gursoy-Ozdemir Y, Vural A, Can A, Topalkara K, Dalkara T. Pericyte contraction induced by oxidative-nitrative stress impairs capillary reflow despite successful opening of an occluded cerebral artery. Nat Med 2009; 15:1031–1037.
22. Chen W, Wang J, Chen J, Chen J, Chen Z. Relationship between changes in the cochlear blood flow and disorder of hearing function induced by blast injury in guinea pigs. Int J Clin Exp Pathol 2013; 6:375–384.
23. Olivetto E, Simoni E, Guaran V, Astolfi L, Martini A. Sensorineural hearing loss and ischemic injury: Development of animal models to assess vascular and oxidative effects. Hear Res 2015; 327:58–68.
24. Shi X. Physiopathology of the cochlear microcirculation. Hear Res 2011; 282:10–24.
25. Ihler F, Sharaf K, Bertlich M, et al. Etanercept prevents decrease of cochlear blood flow dose-dependently caused by tumor necrosis factor alpha. Ann Otol Rhinol Laryngol 2013; 122:468–473.
26. Peppel K, Crawford D, Beutler B. A tumor necrosis factor (TNF) receptor-IgG heavy chain chimeric protein as a bivalent antagonist of TNF activity. J Exp Med 1991; 174:1483–1489.
27. Peppel K, Poltorak A, Melhado I, Jirik F, Beutler B. Expression of a TNF inhibitor in transgenic mice. J Immunol 1993; 151:5699–5703.
28. Ihler F, Pelz S, Coors M, Matthias C, Canis M. Application of a TNF-alpha-inhibitor into the scala tympany after cochlear electrode insertion trauma in guinea pigs: Preliminary audiologic results. Int J Audiol 2014; 53:810–816.
29. Nuttall AL. Velocity of red blood cell flow in capillaries of the guinea pig cochlea. Hear Res 1987; 27:121–128.
30. Zeintl H, Sack FU, Intaglietta M, Messmer K. Computer assisted leukocyte adhesion measurement in intravital microscopy. Int J Microcirc Clin Exp 1989; 8:293–302.
31. Sharaf K, Ihler F, Bertlich M, Reichel C, Berghaus A, Canis M. Tumor Necrosis Factor-induced decrease of cochlear blood flow can be reversed by Etanercept or JTE-013. Otol Neurotol 2016; 37:e203–e208.
32. Kroetsch JT, Bolz S-S. The TNF-alpha/sphingosine-1-phosphate signaling axis drives myogenic responsiveness in heart failure. J Vasc Res 2013; 50:177–185.
33. Ingham NJ, Carlisle F, Pearson S, et al. S1PR2 variants associated with auditory function in humans and endocochlear potential decline in mouse. Sci Rep 2016; 6:28964.
34. Kono M, Belyantseva IA, Skoura A, et al. Deafness and stria vascularis defects in S1P2 receptor-null mice. J Biol Chem 2007; 282:10690–10696.
35. van Eijk IC, Peters MJL, Serne EH, et al. Microvascular function is impaired in ankylosing spondylitis and improves after tumour necrosis factor alpha blockade. Ann Rheum Dis 2009; 68:362–366.
36. Campanati A, Goteri G, Simonetti O, et al. Angiogenesis in psoriatic skin and its modifications after administration of etanercept: Videocapillaroscopic, histological and immunohistochemical evaluation. Int J Immunopathol Pharmacol 2009; 22:371–377.
37. Galarraga B, Belch JJF, Pullar T, Ogston S, Khan F. Clinical improvement in rheumatoid arthritis is associated with healthier microvascular function in patients who respond to antirheumatic therapy. J Rheumatol 2010; 37:521–528.
38. Glacet-Bernard A, Roquet W, Coste A, Peynegre R, Coscas G, Soubrane G. Central retinal vein occlusion and sudden deafness: A possible common pathogenesis. Eur J Ophthalmol 2001; 11:197–199.

Capillary pericytes; Cochlear blood flow; Etanercept; Microcirculation; Pericytes; Tumor necrosis factor

Supplemental Digital Content

Copyright © 2017, Otology & Neurotology, Inc.