Chapter 2. Rationale, Context, Hypotheses & Aims
3.1. Overview
Cervical radiculopathy is a neuropathic condition that induces long-term pain that
radiates from an injured or irritated cervical nerve root to the shoulders and arms (Abbed
and Coumans 2007, Wall and Melzack 1994). Cervical disc herniation is a common
cause of radicular pain in the cervical spine and often involves both compressive and
inflammatory insults to the cervical nerve roots; a bulging disc can physically impinge on
the nerve root and also release materials from its nucleus pulposus, which is
inflammatory to the root (Abbed and Coumans 2007, Caridi et al. 2011, Wall and
Melzack 1994). Rat models imposing compressive or inflammatory nerve root insults
separately have established that while both components of injury can individually induce
sustained mechanical behavioral sensitivity, these stimuli differentially modulate
Winkelstein 2011, Colburn et al. 1999, Hashizume et al. 2000, Hou et al. 2003,
Kawakami et al. 1994, Kawakami et al. 1994, Rothman and Winkelstein 2007). For
example, a transient C7 nerve root compression for 15 minutes in the rat upregulates the
transcription of interleukin-1 beta (IL-1β) and tumor necrosis factor-alpha (TNF-α) in the
spinal cord at 1 hour after compression and activates spinal glia for at least 7 days
(Hubbard and Winkelstein 2005, Rothman et al. 2009, Rothman et al. 2010, Winkelstein
and DeLeo 2002). An inflammatory insult imposed by chromic gut suture does not
induce the same early spinal transcription of IL-1β and TNF-α nor does it lead to the
persistent glial activation that is induced after a root compression (Rothman et al. 2009,
Rothman and Winkelstein 2007, Rothman and Winkelstein 2010). This same nerve root
compression disrupts axonal myelination and promotes phagocytotic macrophage
infiltration at the injured root by two weeks after injury, while an inflammatory insult
only induces macrophage infiltration and to a much less degree than compression (Chang
and Winkelstein 2011). Since both compressive and inflammatory insults to the nerve
root produce indistinguishable mechanical hypersensitivity in the ipsilateral forepaw for
up to 2 weeks (Chang and Winkelstein 2011, Rothman and Winkelstein 2007), but a
compressive insult induces more robust spinal and nerve root pathologies (Chang and
Winkelstein 2011, Rothman and Winkelstein 2007), it is hypothesized that different
cellular cascades are induced by compressive and inflammatory stimuli to produce and
maintain the associated pain responses.
The blood-brain barrier (BBB) and blood-spinal cord barrier (BSCB) are
comprised of endothelial cells and act as physical obstacles between the circulatory
inflammatory molecules and cells under healthy conditions (Abbott et al. 2010, Ballabh
et al. 2004). Despite being remote from the CNS, chronic nerve ligation injury, which
induces sustained pain, disrupts the BSCB for up to 30 days in spinal regions where the
injured afferents synapse in both rats and mice (Beggs et al. 2010, Echeverry et al. 2011,
Gordh et al. 2006). This increased BSCB permeability facilitates the transmission
radioiodine-labeled IL-1β and green fluorescent protein positive (GFP+) bone marrow-
derived monocytes into the spinal cord (Echeverry et al. 2011); both the presence of pro-
inflammatory cytokines and peripheral immune cells centrally can exacerbate
neuroinflammation. Different types of neural injuries can increase spinal microglial
activation through either resident spinal microglial division or through facilitated
infiltration of peripheral monocytes that express markers of microglia activation once
they are in the spinal cord (DeLeo et al. 2004, Rothman et al. 2009, Rutkowski et al.
2002, Stollg and Jander 1999, Sweitzer et al. 2002). A compressive, but not an
inflammatory, insult to the nerve root increases microglia populations in the spinal cord
both early (day 1) and at later times (day 7) after injury (Rothman and Winkelstein 2007).
It is possible that the lack of spinal microglial activation induced by an inflammatory root
insult might be due to its lack of induction of BSCB breakdown preventing peripheral
immune cell transmigration into the spinal cord. Studies in this chapter evaluate whether
compressive or inflammatory insults to the nerve root induce different extents of BSCB
breakdown and whether such barrier breakdown directly mediates pain.
Systemic concentrations of pro- and anti-inflammatory cytokines and chemokines
are upregulated in a variety of neuropathic and autoimmune disorders andinfluence the
Kastin 2007, Pedersen et al. 2015, Sharief and Thompson 1992). Intravenous pro-
inflammatory IL-1β, TNF-α or monocyte chemotactic protein-1 (MCP-1), independently,
increase BSCB permeability in normal naïve rats, which also has been shown to permit
the extravasation of serum proteins into the spinal parenchyma where the BSCB is
compromised (Echeverry et al. 2011, Pan and Kastin 2007, Sharief and Thompson 1992).
In contrast, anti-inflammatory transforming growth factor-beta (TGF-β) and IL-10 rescue
the BSCB breakdown that is induced after painful sciatic nerve ligation (Echeverry et al.
2011); this increased BSCB integrity is also achieved by anti-MCP-1 given intrathecally
over 3 days after ligation (Echeverry et al. 2011). Increased serum levels of some of these
BBB-disrupting cytokines are also increased in patients with painful disc herniation
(Kraychete et al. 2010, Pedersen et al. 2015); in particular, serum concentrations of IL-6,
IL-8 and TNF-α are elevated in those patients experiencing more intense chronic pain
compared to those with “mild” or no pain (Kraychete et al. 2010, Pedersen et al. 2015).
Despite a strong implication of peripheral inflammation in both BSCB breakdown and
pain, the relationship between systemic inflammation at times of increased BSCB
permeability after compressive and/or inflammatory nerve root insults and pain has not
yet been studied.
This chapter summarizes a multi-part study that tests two separate hypotheses.
The first two studies (Section 3.3 & Section 3.4) summarize a subset of the experiments
under Aim 1 and test the hypothesis that painful nerve root injury induces BSCB
breakdown and systemic inflammation facilitating the accumulation of serum molecules
in the spinal parenchyma where the BSCB is compromised (Aim 1a). To test this
BSCB breakdown over time after painful and non-painful nerve root compressions
separately (Section 3.3). The comparative effects of compressive and inflammatory nerve
root insults on BSCB breakdown and pain were also investigated (Section 3.4). BSCB
breakdown was measured in the bilateral spinal cord by immunolabeling for the serum
protein, immunoglobulin G (IgG), which is not typically present in the CNS(Echeverry
et al. 2011, Poduslo et al. 1994), at times early and later after the neural insults (Aim 1a).
To define the relationship between expression of inflammatory serum molecules and pain
at times of BSCB breakdown, the concentration of 23 pro- and anti-inflammatory serum
cytokines and chemokines was correlated to pain severity after a compressive injury
(Section 3.3) or an inflammatory insult (Section 3.4) to the nerve root (Aim 1b). Of the
cytokines found to correlate to pain intensity after a nerve root compression, TNF-α
expression was immunolabeled in the spinal dorsal horn at the time point when BSCB
reaches a maximum after painful and non-painful compression injuries.
The last set of studies in this chapter (Section 3.5) investigates the role of BSCB
breakdown in pain development after a nerve root compression and was designed to test
the hypothesis that blocking BSCB breakdown can inhibit the development of nerve root
compression-induced pain. Activated protein C (APC), a serum protease that strengthens
endothelial tight junctions and effectively reduces vascular leakiness in studies of sepsis
and ischemic stroke (Bernard et al. 2001, Petraglia et al. 2010, Zlokovic and Griffin
2011), was administered intravenously at 1 hour after a painful nerve root injury that
induces BSCB breakdown. The inhibition of BSCB breakdown by APC was confirmed
time to determine if BSCB breakdown contributes to the development of nerve root-
induced pain (Aim 1c).
3.2. Relevant Background
BBB disruption is characteristic of many neurological disorders, including
ischemic stroke, Parkinson’s disease and amyotrophic lateral sclerosis (ALS), and
contributes to the associated neuroinflammation and neurodegeneration within the central
nervous system (Sandoval and Witt 2008, Winkler et al. 2014, Zlokovic 2011). The
neurovasculature of the BBB is comprised of endothelial cells that are trophically
coupled to nearby neurons via glial cells, which together, make up the ‘neurovascular
unit’ and interact closely to maintain inflammatory dysfunction in disease states (Abbott
et al. 2010, Abbott et al. 2006, Hawkins and Davis 2005). The healthy BBB endothelium
is bound together by tight junctions, which inhibit the transmission of serum components
and blood-borne cells into the CNS (Abbott et al. 2010, Ballabh et al. 2004). Disruption
of the BBB permits entrance of neurotoxic factors into the CNS that impair normal
neuronal function and exacerbate inflammation (Webb and Muir 2000, Zlokovic 2008).
Peripheral nerve injuries themselves have also been shown to increase the permeability of
the BSCB at the same spinal level as the injury where the injured afferents synapse
(Beggs et al. 2010, Echeverry et al. 2011, Radu et al. 2013). Nerve injuries also induce
chronic neuropathic pain which is, at least partially, maintained by spinal
neuroinflammation (Echeverry et al. 2011, Rothman and Winkelstein 2007, Watkins et
Further, if BSCB breakdown is related to development of chronic pain, therapeutics
targeting and preventing vascular permeability would be ideal candidates for pain
prevention after neuropathic injury.
Systemic inflammation contributes to spinal neuroinflammation both indirectly
and directly. Pro-inflammatory molecules indirectly influence spinal homeostasis by
stimulating peripheral neuronal receptors (Szelenyi 2001). These primary afferents
synapse centrally in the spinal dorsal horn and become hyperexcitable due to constant
peripheral activation (Szelenyi 2001). Circulating pro-inflammatory cytokines, such as
TNF-α and IL-1β, also activate endothelial cells leading to increased vascular
permeability, which promotes their own entrance into the CNS via a compromised BBB
(Echeverry et al. 2011, Hoffmann et al. 2004, Huber et al. 2001, Pan and Kastin 2007,
Sharief and Thompson 1992). Once in the CNS, pro-inflammatory cytokines and
chemokines can directly interact with and stimulate neurons and glia, which further
accentuate nociceptive pathways by increasing synaptic concentration of excitatory
neurotransmitters and cytokines (DeLeo and Yezierski 2001, Milligan and Watkins
2009). Of note, independently blocking the actions of either spinal TNF-α or IL-1β
attenuates pain after compressive nerve root injury (Rothman et al. 2009), highlighting
the strong influence of these pro-inflammatory cytokines in pain signaling. Systemic and
central inflammation play important roles in pain and are integrated through BBB
permeability(Ballabh et al. 2004, Hawkins and Davis 2005, Huber et al. 2001, Sharief
and Thompson 1992, Szelenyi 2001, Webb and Muir 2000). Yet, it is not known whether
there is a relationship between nerve root-induced pain and systemic levels of
Vascular permeability, including that of the BBB and BSCB, is controlled
through various endothelial pathways. The serine protease, thrombin, most notably
recognized for its role in coagulation, also regulates a variety of endothelial processes
including permeability via its enzymatic activation of cell-bound receptors (Coughlin
2000, Di Cera 2008). Thrombin initiates distinct cellular signaling cascades depending on
the cofactor it binds, and therefore the substrate it activates (Coughlin 2000, Di Cera
2008, Jacques et al. 2000, Komarova et al. 2007, Riewald and Ruf 2005, Xu et al. 2005).
In its unbound state, mammalian thrombin increases vascular permeability by directly
cleaving protease-activated receptor-1 (PAR1) on the endothelial surface (Gandhi et al.
2011, Jacques et al. 2000, Komarova et al. 2007). In contrast, when bound to
thrombomodulin, thrombin activates endothelial-bound protein C into APC, which
stabilizes vascular integrity (Esmon 1993, Fuentes-Prior et al. 2000, Komarova et al.
2007, Xu et al. 2005). Both clinical trials and animal studies have tested APC for its
enhancement of endothelial barriers in sepsis and ischemic stroke (Zlokovic and Griffin
2011); but, its strong anticoagulant effects hamper its clinical safety (Bernard et al. 2001,
Finfer et al. 2008, Marti-Carvajal et al. 2012). For this reason, a major effort in thrombin
mutagenesis and protein engineering has identified peptide domains in thrombin’s
structure that control protein C activation which they can manipulate in order to increase
thrombin’s innate affinity for protein C instead of PAR1 (Dang et al. 1997, Gibbs et al.
1995, Marino et al. 2010). Although APC is not a clinically relevant treatment option for
neural trauma, administering APC after painful nerve root injury to fortify the BSCB
This three-part study was used to test the hypotheses that a painful nerve root
compression induces BSCB breakdown, which facilitates the accumulation of serum
molecules in the spinal parenchyma in regions of breakdown and that the formation of
nerve root-induced pain is inhibited when BSCB breakdown is blocked. Section 3.3
details the methods and results of a characterization study measuring BSCB breakdown
after painful and non-painful nerve root compression by immunolabeling IgG, a serum
protein that is not expressed in the CNS under normal conditions (Poduslo et al. 1994). In
that study, behavioral responses were induced using two previously defined durations of
nerve root compression: 15-minutes to induce sustained pain and 3-minutes to serve as a
loading control in which the nerve root undergoes injury, but pain does not develop
(Nicholson et al. 2012, Rothman et al. 2010). Forepaw mechanical hyperalgesia and
spinal IgG were assessed on days 1 and 7 as time points corresponding to the
development and maintenance of pain, respectively. Expression of serum cytokines and
chemokines were assayed at the time point (day 1) corresponding to BSCB breakdown
using a multiplex assay in order to characterize systemic inflammation and its
relationship to pain.
The second part of this chapter investigates the influence of the type of pain-
inducing nerve root insult (i.e. compressive versus inflammatory) on BSCB breakdown
and behavioral hypersensitivity (Section 3.4). A painful inflammatory insult was imposed
using a rat model of chromic gut suture application to the nerve root (Chang and
Winkelstein 2011, Rothman et al. 2011). Following the same procedures as from the
characterization study (Section 3.3), ipsilateral forepaw mechanical hyperalgesia and
investigate the comparative effects of compressive and inflammatory nerve root insults
on BSCB breakdown (Section 3.4).
The third part of the study defined the role of BSCB breakdown in the
development of pain by blocking BSCB after painful root compression injury (Section
3.5). APC was administered after painful nerve root compression to block BSCB
breakdown, based on its clinical and pre-clinical effectiveness at reducing vascular
disruption (Bernard et al. 2001, Marti-Carvajal et al. 2012). Spinal IgG levels were
measured at day 1 after injury with treatment to confirm its effectiveness and then
behavioral responses were measured at this early time point. The goal of the studies in
this chapter was to identify if BSCB breakdown is induced, and if so, to what
spatiotemporal extent within the spinal cord, after both compressive and inflammatory
nerve root insults. Further, the influence of nerve-root induced BSCB breakdown on pain
development and/or maintenance was investigated. The relationship between serum
inflammatory markers and pain severity was also defined at times of increased BSCB
permeability.