E-Letter responses to:

reports: Lin Jiang, Eric A. Althoff, Fernando R. Clemente, Lindsey Doyle, Daniela Röthlisberger, Alexandre Zanghellini, Jasmine L. Gallaher, Jamie L. Betker, Fujie Tanaka, Carlos F. Barbas, III, Donald Hilvert, Kendall N. Houk, Barry L. Stoddard, and David Baker De Novo Computational Design of Retro-Aldol Enzymes Science 2008; 319: 1387-1391 [Abstract] [Full text] [PDF]

E-Letters: Submit a response to this article

Published E-Letter responses:

Response to D. W. Borhani's E-Letter Comments

David Baker, Eric Althoff, Lin Jiang   (11 August 2008)

Alternative Interpretation of Designed Aldolase Enzymatic Characteristics

David W. Borhani   (11 August 2008)

Response to D. W. Borhani's E-Letter Comments 11 August 2008
David Baker University of Washington, HHMI, Seattle, WA 98195, USA, Eric Althoff, Lin Jiang Respond to this E-Letter: Re: Response to D. W. Borhani's E-Letter Comments	We thank Dr. Borhani for his thoughtful comments on our paper (1). More detailed kinetic analyses currently underway will shed further light on the origins of both the initial burst and the lag phase. With regard to the former, we believe it unlikely that the burst phase arises from a small fraction of active enzyme as the amplitude of the burst is proportional to the substrate rather than the enzyme concentration. With regard to the latter, solution of the rate equations shows that lag phases can arise in multistep reactions when the rates for the elementary steps (imine formation and carbon carbon bond breaking in our case) are within a couple of orders of magnitude. David Baker, Eric Althoff, and Lin Jiang University of Washington, Howard Hughes Medical Institute, Seattle, WA 98195, USA. Reference 1. L. Jiang et al., Science 319, 1387 (2008).
Alternative Interpretation of Designed Aldolase Enzymatic Characteristics 11 August 2008
David W. Borhani, Research Scientist D. E. Shaw Research, LLC Respond to this E-Letter: Re: Alternative Interpretation of Designed Aldolase Enzymatic Characteristics	The work of L. Jiang et al. (Reports, "De novo computational design of retro-aldol enzymes," 7 March 2008, p. 1387) represents an important step forward in synthetic biology and the computational design of functional proteins. I believe, however, that the enzymatic characteristics of some of their designed aldolases may have been misinterpreted. First, constructs RA22 and RA34 exhibited initial-burst kinetics [See fig. 3A; Supporting Online Material, figs. S5E, S5F (1)]. As the authors noted, the size of the burst was quite small; indeed, the observed bursts represent only 1% to 5% of the enzyme amounts. Usually, initial-burst kinetics is observed when a covalent enzyme-substrate intermediate forms, with concomitant release of the first product, much faster than the steady-state enzymatic turnover (e.g., the ultimate rate is limited by second product release or some other slower step) (2). The size of the burst is a direct measure, therefore, of the amount of active enzyme. Thus, if burst kinetics is truly representative of the mechanism of aldolases RA22 and RA34, then only 1% to 5% of these proteins is in an active form, and that active fraction is sensitive to the substrate concentration [See Supporting Online Material, figs. S5E, S5F (1)]. If true, this observation would significantly alter, favorably, the kcat/kuncat ratio. Jiang et al. suggested that "the true enzymatic rate is even faster than the burst phase" [See Supporting Online Material text, p.18 (1)]. I believe there is a more likely interpretation of the experimental data: the initial minor increase in fluorescence intensity (the "burst") represents formation of a weakly fluorescent enzyme-substrate complex. The amount of complex would be approximately proportional to the substrate concentration, as observed. The apparently low amount of complex, relative to enzyme, is due to either a small complex equilibrium constant, the weak fluorescence of the complex (relative to the product 6-methoxy-2-naphthaldehyde), or both. For example, using the apparent KM of the substrate 4-hydroxy-4-(6-methoxy-2-naphthyl)-2-butanone, ~0.5 mM, as the approximate complex equilibrium constant, the fluorescence efficiency of this hypothetical ES complex would be about 6% that of the product. This alternative interpretation may also be consistent with the observation that "interaction between the product and RA22 and RA34 results in a significant decrease in the fluorescent signal." Perhaps close apposition of the substrate and active site residues Trp8, Trp58 (RA34 only), and/or Trp184 is responsible for these effects in RA22 and RA34. Second, constructs RA45 and RA46 exhibited a pronounced lag that Jiang et al. attributed to slow (obligate, covalent intermediate) imine formation (Fig. 3A; Figs. S5C, S5D). If imine formation is slow in the first round of catalysis, however, it will also be slow in all subsequent turnover events, unless the enzyme somehow auto-activates. In other words, slow imine formation would remain the rate-determining step throughout the reaction, and a lag phase would not be observed. It seems more likely that the initial lag phase is due to slow conversion of inactive enzyme to an active form, for example through (de)oligomerization or a slow conformational change. David W. Borhani D. E. Shaw Research, LLC. References 1. L. Jiang et al., Science 319, 1387 (2008) 2. C. Walsh, Enzymatic Reaction Mechanisms (W.H. Freeman & Co. San Francisco, 1979). See especially pp. 67–71.