MadSci Network: Biochemistry
Query:

Re: what is the time taken for a protein molecule to bind to the DNA

Date: Wed Jul 27 09:41:24 2005
Posted By: Dr. James Kranz, Research Scientist
Area of science: Biochemistry
ID: 1121042994.Bc
Message:

Hi Samik,

Thanks for your interesting question; my apologies for the slow response. In thinking more about your question, unfortunately, I really can’t give you as complete an answer as this topic deserves. Since you have listed yourself as being grad level, I’m going to treat your question at a relatively sophisticated level, giving you a starting point to think further about the problem and hope that your own explorations will be relevant to your own problems and interests.

At issue is the variability in mechanism of DNA binding among different types of DNA-binding proteins, including transcription factors, polymerases, nucleases, etc. Some of these proteins are monomeric, some multimeric; some have short half-lives (such as upregulated transcriptional activators) in which signal duration is relatively fleeting (minutes), while some have extremely long half-lives, on the order of the life-time of an organism in cases of “permanent” gene silencing (which, admittedly is controlled by DNA methylation/acetylation). These are examples where the total amount of time that a protein sits on the DNA is quite varied, predictably due to differing affinity for the preferred DNA binding site, and consequently one can predict to observe large differences in the kinetics of DNA association and dissociation).

DNA (and RNA) is fundamentally a mixture of a common phosphodiester backbone chemistry, with base-specific differences for each A, T(U), G, or C in the chain. In terms of the duplex structure of DNA, what a protein “sees” is a mix of a very boring, predictable array of negative charges attached to a ribose sugar, a minor groove side of the base pairs which is almost identical whether a G:C or A:T base pair, and the only distinguishing, non-random feature comprised by the narrow major groove channel, which is just wide enough to accommodate (for example) an alpha helix from a protein. The polymer nature of the phosphodiester backbone forms a cloud of negative charges; DNA-binding proteins are observed to have a net positive charge and so are naturally drawn to the DNA like a moth to a flame. In one case that I’m aware of the in literature, this baseline Coulombic attraction of proteins to DNA due to this intrinsic charge difference has been proposed as an explanation for “faster-than- diffusion” association rates...more aptly described as an accelerated diffusion, perhaps.

The commonalities are smaller than the differences in mechanism of action among RNA-binding proteins and DNA-binding proteins, but they do generally share the common feature of exhibiting specific binding to a preferred binding site as well as weaker non-specific binding to a DNA or RNA scaffold. My own personal research history is stronger in the area of protein:RNA binding, so we’ll borrow a brief example to illustrate the point. hnRNP proteins share a common RNA-binding motif with many other proteins, including the U1A protein which comprises a part of the U1 snRNP in RNA splicing; the U1A protein has 0.01-1 nM affinity for its preferred substrate and only ~ 10 uM affinity for non-specific sequence, while hnRNP proteins generally only have ~1 uM affinity for a preferred sequence with 10-fold weaker binding to non-specific sequences. In this example, the on- rate of U1A for it’s preferred substrate is diffusion limited, with off- rates on the time course of minutes; conversely, both on- and off-rates for hnRNP proteins are fast, occurring on the order of milliseconds. These differences reflect a difference in function, with U1A protein serving as a “permanent” anchor on U1 snRNA while hnRNP proteins act as single-stranded, non-specific binding proteins, protecting all forms of “naked” RNA from degradation. So to must the kinetics of DNA- association reflect a real difference in biological function; the short answer, as always, is “it depends”...it depends on the particular example.

The other complication to addressing your question concerns the mechanism through which DNA-binding proteins are thought to find a preferred, specific binding site, given a common feature of non-specific binding to a random DNA sequence. More so in the case of DNA binding proteins than for RNA binding proteins, virtually all DNA-binding proteins are observed to bind to random (homo- and heteropolymer) sequences of DNA. Briefly, a “specific” binding site is characterized by direct, Hydrogen-bonding to donors and acceptors of major groove base pairs, while non-specific binding tends to lack these direct contacts to the DNA bases. In general, the mechanism of DNA-binding involves association with the DNA backbone, normally in the form of ionic pairing with the charged phosphates

The REAL complication comes in describing how different proteins achieve this approach and capture of a long DNA strand, containing a short binding site. Thinking about the disparity in size among the relevant players, proteins are always small in comparison to the DNA or RNA that they straddle. Even the ribosome, which is huge (>500 kDA, and about 50% is RNA) for a protein, is small compared to an average mRNA (assuming ~250 g/mol per nucleotide, so ~750 g/mol per codon, we can readily estimate ~100kDa worth of RNA for a small ~100 amino acid protein); chromosomal DNA, which is 10^6 base pairs in the smallest bacteria and 10^9 in humans, presents a sea of non-specific sequences for a single preferred biding site. You should take the time to read a few reviews on the subject, in particular one from Stephen E Halford and John F Marko:

Halford & Marko (2004) How do site-specific DNA-binding proteins find their targets? Nuc. Acids Res. 32: 3040-3052

The essential discussion concerns the different likely mechanisms through which proteins locate their specific target binding site rapidly, given that these sites comprise a “minute fraction of the cellular DNA”. There are essentially three models to consider (and it appears each of these is relevant to biological function on a case-by-case basis). The first model, DNA “sliding”, involves non-specific binding to DNA followed by translocation along the DNA, without dissociating, until the specific binding site is found. The second model, “hoping”, involves rapid biding and dissociating, or sampling of the DNA, until the preferred binding site is found. The third model, “intersegmental transfer”, involves transient binding of looped DNA, transferring from one site to another distal site; this model generally requires a dimeric protein (with minimally two DNA binding so that one protein unit/dimer can bind to two pieces of DNA simultaneously). Once you appreciate the differences in these models, you’ll understand why it is difficult to answer your question concerning binding kinetics in a simple way.

One more word about rates in general; most biological processes are fast. Active site chemistry, molecular diffusion and molecular reorientation (tumbling) generally occurs on the picosecond to nanosecond time scale, binding events and large conformational changes generally occur on the nanosecond to microsecond time scale, complex reactions, protein folding and large-scale translations occur in the sub-microsecond to millisecond time scale, dissociation kinetics in the millisecond to second time scale, thermodynamically unfavored transitions (large energy barrier events) such as praline isomerization can be slower still. Ignoring system-specific differences and exceptions, most binding events are relatively fast, and most off-rates are much slower; the magnitude of the binding energy should tell you something about what the relative rates must be.

I hope this gives you a starting point for your own interests. Thanks for the challenging question.

Regards,
Dr. James Kranz


Current Queue | Current Queue for Biochemistry | Biochemistry archives

Try the links in the MadSci Library for more information on Biochemistry.



MadSci Home | Information | Search | Random Knowledge Generator | MadSci Archives | Mad Library | MAD Labs | MAD FAQs | Ask a ? | Join Us! | Help Support MadSci


MadSci Network, webadmin@madsci.org
© 1995-2005. All rights reserved.