Gary B. Rollman,
Emeritus Professor of Psychology,
University of Western Ontario
(In addition to links below, see weekly archives in the right column)
Friday, July 15, 2011
Specificity Versus Patterning Theory: Continuing the Debate - Allan Basbaum | Pain Research Forum
Does the brain process and interpret innocuous and noxious stimuli by "reading" a pattern of activity across multimodal lines of activity, or are there specific, labeled lines that carry functionally distinct modalities from the periphery to the spinal cord and then rostrally in the neuraxis?
The specificity camp has its origins in the nineteenth-century studies of German neurophysiologists who concluded that there are modality-specific spots on the skin (e.g., touch, cold, etc.), and that a percept is generated by activation of specific neuronal pathways in the periphery and CNS (for reviews, see Craig, 2003; Ma, 2010; and Perl, 2007). This concept was originally articulated by Müller (Müller, 1840), who proposed that the stimulus did not even determine the perception, but rather that connectivity of the afferent and its ultimate projection site in the brain were critical.
The pattern theory, by contrast, proposed that afferent fibers respond to a host of stimulus modalities, and that the ultimate perception depends on the brain's deciphering and interpretation of the patterns of activity across the different nerve fibers. The late Patrick Wall, almost 30 years after coauthoring the Gate Control Theory of Pain in 1965 with Ronald Melzack (Melzack and Wall, 1965), wrote that "…specificity theory has failed to generate any explanation for clinical pains. Worse yet…it has encouraged ineffective, often counterproductive, surgical attempts to destroy the cells or their axons" (Wall, 1996). The poster boy/dartboard for the specificity camp, of course, is Descartes' little boy whose "pain pathway" runs from his foot to a pain center in his brain.
It is of interest that the focus of Melzack and Wall's seminal 1962 article "On the nature of cutaneous sensory mechanisms," which preceded the Gate Control Theory paper by three years, was on the processing of non-noxious thermal and mechanical stimuli, as compared to pain-producing stimuli. The paper dealt much less so, if at all, with the question of the processing of different modalities of noxious/painful stimuli (i.e., heat, cold, mechanical, and chemical pain) (Melzack and Wall, 1962). Pat Wall was, of course, strongly influenced by his discovery, with Lorne Mendell, of the wide dynamic range (WDR) dorsal horn neuron (Mendell and Wall, 1965). The WDR neuron clearly responds to both innocuous and noxious stimulation, which argued against a specific Cartesian "pain" pathway. However, subsequent demonstrations by Ed Perl and colleagues of primary afferents (nociceptors) and lamina I dorsal horn neurons that respond only to noxious stimulation, at least in uninjured animals (Bessou et al., 1971; Christensen and Perl, 1970) provided fodder for the labeled line advocates.
So where are we today? As a student of both Melzack and Wall, I grew up with a firm belief in pattern theory. However, results from our recent studies have encouraged me to reexamine this basic tenet. Just as the ability to record from single fibers made possible the identification of afferents that respond rather exclusively to noxious stimulation, so the molecular revolution has revealed a detailed subclassification of the nociceptors. In fact, nociceptors are remarkably heterogeneous; different subsets express channels that are responsive to different noxious stimulus modalities (e.g., heat [TRPV1], cold [TRPM8], mustard oil [TRPA1], etc.) (Basbaum et al., 2009; Julius and Basbaum, 2001). Of course, there is some overlap of these channels, and there is no question that many of the nociceptors are polymodal, responding to thermal as well as mechanical stimulation. Nevertheless, our studies argue strongly for a modality-specific contribution of subsets of primary afferents to the presumptive pain-generated behaviors evoked by the particular modality, (e.g., heat, cold, mechanical) (Cavanaugh et al., 2009; Scherrer et al., 2009).
Although I have written (in a syllabus for an IASP Refresher course) that "there are no labeled lines….", I now believe that there is behaviorally relevant specificity, at least at the level of the primary afferent. To what extent the behaviorally relevant (pain, and likely itch) information generated by primary afferents is also manifest at the level of circuits in the spinal cord and at higher levels of neuraxis is the critical unanswered question. Yes, there are nociceptive-specific neurons in the dorsal horn, but the WDR neuron cannot be ignored (Price et al., 2003). Indeed, the relative contribution of these two neurons is worthy of continued discussion in this forum.
In their 1962 article, Melzack and Wall proposed that, "A satisfactory theory of somesthesis must be able to provide answers to two essential problems: 1) What is the nature of the information that is sent to the central nervous system when the skin is stimulated, and 2) How do the central cells select or abstract from this information to provide the many different qualities of our sensory experience?" We have come a long way to answering the first question, butclearly, we need more information as to the specificity versus patterning question at the level of CNS circuitry, so that answers to the second question can be generated.
These questions are not merely of interest to the basic scientist, but are relevant to the development of approaches to the clinical management of pain. Are there clinical pain conditions that arise from activity in subsets of nociceptors, and if so, can drugs be developed to block selectively the contribution of those afferents? For example, TRPV1 is generally associated with noxious heat transduction and heat hypersensitivity, but clearly that is not its function in a visceral afferent that innervates the pancreas. Thus, a drug that blocks TRPV1-expressing afferents selectively may have great utility, beyond regulating heat-pain sensibility (which is clearly not a major clinical concern).
The question of specificity is also relevant to ablative procedures. Because attempts in animals are now made to ablate chemically subsets of spinal cord neurons (e.g., with substance P-conjugated to saporin; Nichols et al., 1999), the controversy, unquestionably, has significant clinical implications. Does the preclinical development of these techniques mean that we are poised to go beyond what Pat Wall, as noted above, referred to as "ineffective, often counterproductive, surgical attempts to destroy the cells or their axons" (epitomized by anterolateral cordotomy), or are the new approaches built upon a misunderstanding of the way pain is generated? Should we be concerned that reducing information flow in a specificity-based "pain transmission network" can contribute to the development of central pain syndromes (as can occur post-stroke)?
Finally, and very importantly, this author certainly appreciates that this contemporary perspective on the question of specificity versus patterning relates more to the processing of nociceptive messages, and much less so to the sensory experience/perception of pain. The latter is clearly influenced and in some cases dominated by emotional and cognitive factors. To what extent specificity or patterning or some hybrid model integrates with these factors is not at all clear, and will not be determined by studies directed only at the primary afferent or spinal cord circuits. Studies that integrate an analysis of the pathways through which inputs are transmitted from the cord to the brain, with others directed at identifying where and how emotional/affective and cognitive factors are processed, will require combining psychophysical studies with novel imaging methods that can monitor the process of information transfer from the spinal cord to the brain.
Remember that the original discussion of specificity versus patterning was not a neurophysiological one. Rather, it was a perceptual one: How does the brain generate a pain percept? It is certainly possible that convergence of "specific" inputs at the level of the spinal cord or higher in the neuraxis generates an integrated pattern of activity that is read by the brain, the product of which is the ultimate percept.
Although many philosophical questions may ultimately be answered by neurobiology—the nature of self-knowledge and free will, for example—generally the two disciplines live in parallel separate universes. My first exposure in the late 1970s to the Specificity versus Pattern Theory debate, repeated endlessly over too many pints of ale at the Jeremy Bentham pub near University College London, was from Pat Wall. He was passionate in his hostility to Specificity, based I think on his philosophical certainty that the Pattern theory was much more intellectually challenging and interesting than "mere fixed telephone lines," and therefore it just had to be correct, a question of neuroaesthetics. However, in the end, facts are our currency and the accumulated data from the last 30 years demonstrate quite conclusively that at the primary afferent level, there is clear specificity. Sensory neuron subtypes display exquisite selectivity in deconvoluting the external work into defined neural representations by virtue of their specialized transduction processes, extracting information about particular aspects of the sensory environment, its intensity, nature, place, and time. Individual afferents capture enough information from defined stimuli to tempt us to conflate their firing patterns with the sensation generated, leading us into calling them warm, tickle, itch, or pain fibers. However, as Basbaum points out in describing his conversion from the Pattern Theory to the Specificity School, the way in which the CNS represents the sensory information at a systems level and the way in which it is read from neural circuitry as a sensation remains almost a closed book. Is the exquisite specificity engineered in the periphery maintained in the CNS, or is perception the consequence of an interrogation of complex multidimensional information across space and time, the translation of a pattern? Here, due to our lack of data, we can remain, for the moment, philosophical. However, the capacity to silence or activate designated subsets of central neurons, to record activity simultaneously from many neurons in vivo, and the new field of connectomics that will reveal the synaptic wiring diagram of the brain, means that it is only a matter of time before the answer will be revealed. Take your bets. I know mine.