Arthur Karlin

Principal Investigator

Dr. Arthur Karlin is a Higgins Professor of Biochemistry, Molecular Biochemistry and Molecular Biophysics, Physiology and Cellular Biophysics, and Neurology at Columbia University, and the Director of the University's Center for Molecular Recognition.

Dr. Karlin's Work

Dr. Karlin has spent the last 40 years researching acetylcholine receptors. In the process of his work, he initiated the biochemical study of membrane receptors and demonstrated that the nicotinic receptor for the neurotransmitter acetylcholine is a protein. As part of this discovery, he also identified the acetylcholine-binding subunit, isolated the intact receptor, determined its molecular weight, identified all of its subunits and their stoichiometry and arrangement, and determined the molecular properties of the ACh-receptor binding sites, channel, and gate.

Additionally, Dr. Karlin developed the substituted-cysteine-accessibility method (SCAM) and its applications to the identification of the residues lining pores and binding sites, the detection of changes in the accessibility of these residues with changes in functional state, the location of pore gates, and the mapping of the electrostatic potential in pores and binding sites.

Over the last 8 years, Dr. Karlin has worked on the large conductance, voltage- and calcium-activated potassium channel (BK), including investigating its structure and function. In particular, in collaboration with Steven Marx, he determined where its unique seventh transmembrane helix (S0) interacts with the other six helices, and demonstrated that the modulatory beta1 subunit exerts its influence in part through contact with S0. Concurrently during the last 3 years, Dr. Karlin developed a unique mathematical model of the control of membrane potential and Ca2+ concentration in an arterial smooth muscle cell, involving 37 channels, transporters, receptors, and enzymes and five separate but communicating compartments and microdomains for Ca2+. Dr. Karlin will extend this model to include endothelial cells.

A bibliography of Dr. Karlin's work is available here.


Steven O. Marx, M.D.
Professor of Medicine (in Pharmacology)
Columbia University

Guoxia Liu, Ph.D.
Research Scientist, Division of Cardiology, Department of Medicine
Columbia University

Sergey I. Zakharov, Ph.D.
Research Scientist, Division of Cardiology, Department of Medicine
Columbia University

Positions and Honors

Academic Positions:

  • 1962-1963: Research Assistant, College of Physicians & Surgeons, Columbia University
  • 1963-1964: Research Associate, College of Physicians & Surgeons, Columbia University
  • 1965-1969: Assistant Professor of Neurology, P & S, Columbia University
  • 1969-1974: Associate Professor of Physiology in Neurology, P & S, Columbia University
  • 1974-1978: Associate Professor of Neurochemistry, P & S, Columbia University
  • 1978-present: Professor of Biochemistry and Molecular Biophysics, and Neurology
  • 1989-present: Higgins Professor of Biochemistry and Molecular Biophysics, and Neurology, and Director of the Center for Molecular Recognition, P & S, Columbia University
  • 1991- present: Professor of Physiology and Cellular Biophysics, P & S, Columbia University.

Other Experience and Professional Memberships:

  • 1975: Chairman, Gordon Research Conference on Molecular Pharmacology
  • 1978: NIH Pharmacology A (AHR) Study Section
  • 1980: NIH Neurology B Study Section
  • 1982: NIH Special Study Section
  • 1982: Visiting Committee on Neurobiology, Salk Institute
  • 1983: NIMH Neurosciences Program Evaluation Consultant
  • 1985-1989: Director, Summer Neurobiology Course, Marine Biological Laboratory
  • 1988: Visiting Committee on the Department of Biochemistry, Brandeis University
  • 1995-1999: Fachbeirat, Max Planck Inst. fur medizinische Forschung in Heidelberg
  • 2007- 2010: Chairman, Section 23 – Physiology and Pharmacology, National Academy of Sciences
  • 2009: Ad hoc member of BPNS NIH review panel

Editorial Boards

  • 1972-1983: Molecular Pharmacology
  • 1979-1984, 1987-1991: Journal of Biological Chemistry
  • 1986-1998: Proteins
  • 1990-1995: Journal of Neuro­science
  • 1996-2000: Neuron.
  • 2009 - Present: Proceedings of the National Academy of Sciences


  • 1975: Lucy G. Moses Prize in Basic Neurology; 1979
  • 1984: Grass Traveling Scien­tist, Vanderbilt University
  • 1984: J. H. Quastel Visiting Professor, McGill University
  • 1985: John C. Krantz, Jr. Lectureship in Pharmacology and Experimental Therapeutics, U. Md.
  • 1985: The Louis and Bert Freedman Foundation Award for Research in Biochemistry, NYAS
  • 1987: Javits Neuro­science Investigator Award
  • 1989: Fellow of the American Association for the Advance­ment of Science
  • 1989: Stevens Triennial Prize, College of Physicians and Surgeons
  • 1989: Harvey Society Lecturer
  • 1990: Dean's Distinguished Lecturer in the Basic Sci­ences
  • 1994: Fellow of the American Academy of Arts and Sciences
  • 1999: Member of National Academy of Sciences
  • 2004: Noun Shavit Lecturer, Ben Gurion University

Contributions to Science

The first molecular characterization of a membrane receptor and of a receptor ligand binding site: In graduate school I was attracted by the concept of receptor, which motivated my thesis work on the regulation by neurohypophyseal hormones of water permeability. I went to David Nachmansohn at Columbia as a postdoctoral to work on the acetylcholine (ACh) receptor. There was little appreciation at the time that receptors, channels, and transporters were intrinsic membrane proteins or even what that meant. There was still debate about whether ACh receptor (a pharmacological concept) was a protein. I tried to get a handle on the receptor.  With Eva Bartels, who was adept at recording the ACh-induced depolarization of isolated  electroplax from Electophorus (electric eel), I tried to characterize the receptor by perturbing it with various protein modifying reagents. The most promising was the reducing agent, dithiothreitol (DTT). A few minutes exposure of the electroplax to DTT inhibited the receptor. This could not be reversed by washing but could be completely reversed by an oxidizing agent like DTNB. Most importantly, this reversal could be blocked by a SH alkylating agent like N-ethylmaleimide. We published this in 1966, which was the first evidence that this receptor, because it had disulfide bond, was likely to be a protein. On the chance that the reducible disulfide was close to the ACh binding site, I designed a quaternary ammonium benzyl maleimide (MBTA) to affinity label the CySH generated by DTT reduction. This reacted with the receptor 1000 times faster than did the uncharged N-ethylmaleimide. I was inspired to try this by Irwin Wilson’s development of PAM, an affinity reactivator of acetylcholinesterase covalently blocked by “nerve gases” and by S.J. Singer’s affinity labeling of antibody binding sites. Subsequently, Israel Silman and I designed other affinity reagents (e.g. bromoacetylcholine) that permanently activated the receptor – these were the first tethered agonists. I observed that shorter affinity labels permanently activated the receptor, and longer affinity labels permanently inhibited the receptor. The implication was that reversible agonists (e.g., ACh) promoted a contracted binding site and that reversible antagonists (e.g. curare) maintained an expanded binding site [1a]. With [3H]MBTA, we affinity labeled reduced ACh receptor in intact eel electroplax, extracted the receptor in SDS, electrophoresed it, and obtained a single radioactive band of ~40 kDa [1b]. This was the first characterization of any membrane receptor as a protein. We purified the detergent-solubilized receptor from Electrophorus electric tissue and from Torpedo electric tissue [1c]. The latter preparation gave four distinct bands: 40 (α), 48 (β), 58 (γ), and 64 (δ) kDa, and only α was affinity labeled [1c]. After Numa cloned and sequenced α, we first identified the affinity labeled residues in the ACh binding site as Cys192 and Cys193 in 1983 and latter showed that these formed a highly unusual disulfide between adjacent Cys [1d]. Cys192 and Cys193 were only found in ACh receptor α subunits. Two additional residues not at the binding site, Cys128 and Cys142, also formed a disulfide [1d]; Cys128 and Cys142 are conserved in all subunits and all species of the ACh receptor and in almost all subunits of all other homologous receptors in what became known as the Cys-loop receptors.

1a. Karlin, A. 1969. Chemical modification of the active site of the acetyl­choline receptor. J. Gen. Physiol. 54:245s-264s.

1b. Reiter, M.J., Cowburn, D.A., Prives, J.M. and Karlin, A. 1972. Affinity labeling of the acetylcholine receptor in the electroplax: Electrophoretic separation in sodium dodecyl sulfate. Proc. Natl. Acad. Sci. U.S.A. 69:1168-1172.

1c. Weill, C.L., McNamee, M.G. and Karlin, A. 1974. Affinity labeling of purified acetylcholine receptor from Torpedo californica. Biochem. Biophys. Res. Commun. 61:997-1003.

1d. Kao, P.N., and Karlin, A. 1986. Acetylcholine receptor binding site contains disulfide crosslink between adjacent half-cystinyl residues.  J. Biol. Chem. 261:8085-8088.

Physical properties of the muscle-type nicotinic ACh receptor:
By analytical ultracentrifugation of pureTorpedo ACh receptor in non-ionic detergent, Jackie Reynolds and I determined that the molecular weight of the intact receptor was 250 kDa [2a]. Our preparation had 2 moles of binding sites per 250,000 g protein. Thus, there were two α subunits per complex. The only possible stoichiometry that would add to 250,000 was α2βγδ. The ‘Cys-loop receptors’ became the ‘pentameric ionophoric receptors’. In fact, Torpedo ACh receptor is mostly a disulfide-crosslinked dimer of 500 kDa. This disulfide is between δ subunits [2b], which we took advantageous of to determine by single-particle electron microscopy the order of the five subunits around the central pore [2c]. Our order was controversial, and the competing incorrect order continued to appear for some time in text books. What was of consequence was the insight that the first ACh binding site was formed at the interface between the first α and γ and the second ACh binding site was formed between α and δ [2d]. None was formed between β and δ. We had observed much earlier that the two ACh binding sites behaved differently, and we now had a structural basis for this.

2a. Reynolds, J.A. and Karlin, A. 1978. Molecular weight in detergent solu­tion of acetylcholine receptor from Torpedo californica. Biochem. 17:2035-2038.

2b. Hamilton, S.L., McLaughlin, M. and Karlin, A. 1979. Formation of disul­fide-linked oligomers of acetylcholine receptor in membrane from Torpedo electric tissue. Biochem. 18:155-163.

2c. Karlin, A., Holtzman, E., Yodh, N., Lobel, P., Wall, J. and Hainfeld, J. 1983. The arrangement of the subunits of the acetylcholine receptor of Torpedo californica. J. Biol. Chem. 258:6678-6681.

2d. Czajkowski, C., and Karlin, A.  1995. Structure of the nicotinic recep­tor acetylcholine binding site: Identification of acidic residues in the delta subunit within 0.9 nm of the alpha subunit binding site disulfide. J. Biol. Chem. 270:3160-3164.

Substituted-cysteine-accessibility method (SCAM):
In the 1980s, several laboratories including mine had used photolabeling or site-directed mutagenesis to identify residues lining the pore of the ACh receptor. The outcome was that three residues in the second membrane-spanning helix, M2, and two more in its intracellular flank were likely to face the channel in each of the five subunits. I conceived of a method to identify most of the residues facing a pore, which I called the substituted-cysteine accessibility method (SCAM). I designed reagents based on methanethiosulfonate (MTS) reactivity that had head groups of different size or charge. Initially, these were MTSethylamine (MTSEA), MTSethyltrimethylammonium (MTSET), and MTSethylsulfonate (MTSES). The idea was to mutate  to Cys, one at a time, all residues in each potential channel-lining segment. If the mutant showed near-normal function, then we determined whether the MTS reagents had irreversible effects on channel conductance; i.e, reacted with the Cys. David Stauffer synthesized these compounds, and Myles Akabas tested them on Cys-substitution mutants in the ACh receptor αM2 helix [3a]  In the ACh receptor, we identified all the residues that lined the channel in α and β, M1 and M2, confirming the earlier results and identifying additional channel-lining residues. Many Cys showed state-dependent activity. Stauffer and I determined the rates of reaction of differently charged MTS reagents with the ACh binding site Cys, from which we could infer the electrostatic potential at the site. Javitch and I applied SCAM to determine the identity of the residues in the dopamine binding site in D2 receptors [3b]. Many other channels were probed, for example the NMDA channel [3c]. Recently, Lingle applied SCAM to S6 in the BK channel. We used SCAM to locate the gate in the ACh receptor, which is different in the closed state and in the desensitized state [3d].

3a. Akabas, M. H., Stauffer, D. A., Xu, M., and Karlin, A.  1992.  Acetyl­choline receptor channel structure probed in cysteine-substitution mutants.  Science 258:307-310.

3b. Javitch, J., Fu, D., Chen, J., and Karlin, A. 1995. Mapping the binding-site crevice of the dopamine D2 receptor by the substituted-cysteine acces­sibility method.  Neuron 14:825-831.

3c. Kuner, T., Wollmuth, L.P., Karlin, A., Seeburg, P., and Sakmann, B. 1996. Structure of the NMDA receptor channel M2 segment inferred from the accessibility of substituted cysteines. Neuron 17:343-352.

3d. Wilson, G.G., and Karlin, A. 2001. Acetylcholine receptor channel structure in the resting, open, and desensitized states probed with the substituted-cysteine-accessibility method. Proc. Natl. Acad. Sci. USA 98: 1241-1248.

Characterizations of BK α and β subunits and of their complexes:
Steven Marx and I have been working together on the BK channel since 2006. Our approach combines biochemistry and electrophysiology. We determine the susceptibility of substituted Cys in pairs to disulfide crosslinking and infer from that their relative proximities. We also determine the functional effects of crosslinking. From these two kinds of evidence and from molecular modeling based on the Kv1.2/2.1 structure, we have inferred which segments of the structure are neighbors and whether or not these segments need to move relative to one another in the transitions between functional states. We determined that S0 is close to S3-S4, which could be significant in the uniquely far-rightward shifted G-V curve of BK compared to other voltage-gated K channels. After we extended the Cys substitution into the first helical turn in the membrane, we were able to locate S0 as close to S3 and S4 but outside the S1-S4 bundle [4a]. We also located the extracellular ends of β1 TM1 and TM2. These are next to each other in the cleft with TM1 next to S1 and S2, in one α, and TM2 next to S0 in the adjacent α [4a].  The rate of crosslinking of S0 to the short loop between S3-S4 was not sensitive to whether BK was activated or deactivated, but crosslinking was faster in the presence of β1, as if TM1 and TM2 were filling the cleft between αs and pushing S0 close to S3-S4 [4b].  Examination of the disposition of β2 and β3a in complexes with α revealed that mouse β3a, which has an unpaired extracellular Cys forms disulfide with either αCys14 or αCys141, ordinarily crosslinked to each other. This lead us to determine directly the crosslinking pattern of the four conserved Cys in the β1 extracellular loop and, by homology, the disulfide crosslinking pattern in all of the β types [4c]. In the intracellular domain, Hoshi had determined that residues in the S0-S1 loop formed a Mg2+ binding site with residues in the RCK domains and Magleby had determined that potentiation by Ca2+ was sensitive to the length of the linker between the S6 helix and the C-terminal RCK domains. We determined the relative positions of the flank of S0 and the flanks and loops of S1-S6 [4d].  In the absence of β1, S0 was close to the S2-S3 loop. In the presence of β1, however, S0 was displaced away from S3 by TM2.

4a. Liu, G., Niu, X., Wu, R.S., Chudasama, N., Yao, Y., Jin, X., Weinberg, R., Zakharov, S.I., Motoike, H., Marx, S.O., and Karlin, A. (2010)  Location of modulatory b subunits in BK potassium channels. J. Gen. Physiol. 135:449-459; PMCID:  PMC2860586

4b. Niu X, Liu G, Wu RS, Chudasama N, Zakharov SI, et al. (2013) Orientations and proximities of the extracellular ends of transmembrane helices S0 and S4 in open and closed BK potassium channels. PLoS ONE 8(3): e58335. doi:10.1371/journal.pone.0058335 PMCID: PMC3589268

4c. Wu RS, Liu G, Zakharov SI, Chudasama N, Motoike H, Karlin A, Marx SO. (2013 ) Positions of β2 and β3 subunits in the large-conductance calcium- and voltage-activated BK potassium channel. J Gen Physiol. 141:105-17. PMCID:PMC3536527

4d. Liu G, Zakharov SI, Yao Y, Marx SO, and Karlin, A (2015) Positions of  the cytoplasmic end of  BK alpha S0 helix relative to S1-S6 and of beta1 TM1 and TM2 relative to S0-S6. J Gen Physiol 145:185-199. doi:10.1085/jgp.201411337

Modeling kinetic mechanisms of proteins and cells: In my thesis, I modeled the dependence of water transport in toad bladder on the concentration of arginine vasopressin and on time. The dependence on concentration was fit with a Hill equation with an exponent >1. Furthermore, I observed that the response to ACh of the ACh receptor in isolated electroplax was fit with a Hill coefficient of 1.8; after reaction of the receptor with DTT, however, the concentration response curve was fit with a Hill coefficient of 1.1. Thus, I was primed when I read the papers of Monod, Wyman, Jacob and Changeux on ‘allosteric’ interactions. They focused on cooperative binding and its effect on enzyme reactivity. I realized that the theory as it pertained to the equilibrium between the T and the R state could be useful to model receptor responses. I expanded their model to include binding to both the R and T states and to include competitive antagonists and allosteric antagonists and co-agonists [5a].  I used the model to fit the data I had obtained on the response of the ACh receptor. The model, which has relatively few parameters and is useful for fitting the responses of receptors, depends on the assumption of a symmetrical complex that undergoes only synchronous (‘coherent’) transitions of its subunits between just two states. This assumption cannot be correct for the essentially asymmetrical muscle-type ACh receptor with its α2βγδ composition and ACh binding sites with different properties. Like most models, it is a useful simplification until it is not. My colleague, Sam Silverstein, asked me to help him fit data he had obtained on the killing of bacteria by neutrophils. I modeled this as the competition between a first-order reaction (bacterial growth) and a second-order reaction (bacterial killing by neutrophils).  The model provided a useful parameter, the critical neutrophil concentration, below which bacterial concentration increased and above which it decreased, independent of the initial bacterial concentration (5b, 5c). Recently, I have been modeling the control of membrane potential and of sarcoplasmic Ca2+ in arterial smooth muscle. I was trying to understand the role of BK and the mechanisms whereby Ca2+ acts as a signal both for contraction and relaxation. Conceptually, some of this had been worked out by Mark Nelson and his many students, and clearly Ca2+- microdomains, which have been widely discussed in the literature, were involved, but many aspects were undefined. Nelson had published data on the autonomous response of arterial smooth muscle to stretch caused by intravascular pressure. I developed a kinetic model of a smooth muscle cell involving 37 protein components and five communicating compartments, including two microdomains [5d]. I fit this model to Nelson’s determination of membrane potential and global sarcoplasmic Ca2+ as a function of pressure in isolated sections of cerebral arteries. The model also simulated Nelson’s data on the effects of blocking BK, ryanodine receptor, and L-type Ca channels.

5a. Karlin, A. (1967) On the application of "a plausible model" of allosteric proteins to the receptor for acetylcholine. J. Theoret. Biol. 16:306-320.

5b. Li, Y., Karlin, A., Loike, J.D., and Silverstein, S. C. (2002) A critical concentration of neutrophils is required for effective bacterial killing in suspension. Proc. Natl. Acad. Sci. USA 99: 8289-8294.

5c. Li, Y., Karlin, A., Loike, J.D., and Silverstein, S. C. (2004) A critical concentration of neutrophils required to block growth of Staphylococcus epidermidis in fibrin gels and of E. coli in rabbit dermis. J. Exp. Med.200:613-622.

5d. Karlin A (2015) Membrane potential and Ca2+ concentration dependence on pressure and vasoactive agents in arterial smooth muscle: A model. J. Gen. Physiol. 146:79–96