Orthology, sequence similarity, and conservation of function

Question

[Dunois]

What exactly is the relationship between sequence similarity, orthology, and conservation of function?

It is assumed that similar sequences perform similar functions. Is the converse also true?

[GenoMax]

This will be a useful read: https://www.ncbi.nlm.nih.gov/books/NBK20255/

[Dunois]

Many thanks for the link to this book!!

I am still reading, and I have a lot of information to digest. It also addresses some of the aspects the answers by the others have covered. The concept of "functional orthologs" (as the book calls it) is what I am interested in, and I'll dig into this further.

(I think I will try and read this book cover to cover eventually),

[lieven.sterck]

some more info in this thread as well (not 100% related but useful nonetheless) : Orthologs vs PFAM

Answer 1

[Carlo Yague]

It all comes down to evolution. True orthologous genes often have similar sequences and similar functions because they come from the same ancestor gene/proteins. Of course if the evolutionary distance between the orthologs is high, then the sequence divergence will be higher and it becomes more lilely that one of the two species have evolved new function(s) or lost ancestral function(s) for the gene.

That being said, I have to disagree with your thesis:

It is assumed that similar sequences perform similar functions.

An common example when this is not necessarily true is when there are paralog genes. Paralogs evolve when a gene becomes duplicated in a particular species. They have similar sequences, but often, they end up evolving new functions. This happen because the important/essential ancestral function is being taken care of by one of the paralog, lifting evolutive pressure from the other one, which becomes "free" to mutate and evolve.

Is the converse also true?

Even if we forget about paralogs, the converse of the above hypothesis would not be true. This is typically what is called "evolutive convergence". Convergence is the independent evolution of similar functions in different species when the gene that give that function was not rooted in their common ancestor (like wings in bats and birds). So similar function, but not necessarily similar sequence.

Answer 2

[Mensur Dlakic]

It is assumed that similar sequences perform similar functions. Is the converse also true?

Fairly similar sequences are likely to perform similar functions. Let's call those similar proteins members of protein families. Somewhat similar sequences will perform somewhat similar functions, and those members would belong to protein superfamilies. A while ago I described a superfamily of phosphoesterases that are very different in terms of sequence, yet they all cleave phosphate esters. When I tell you that some of them are phosphatases and others are nucleases, you may think they are nothing alike. Yet both single-phosphate removal from any substrate and a breakdown of a phosphodiester backbone in DNA/RNA are chemically identical reactions, so these enzymes perform somewhat similar functions even though their sequences often share less than 20% identity. There is a whole spectrum of what similar means, and the meaning will be different when applied to the chemical nature of reactions vs chemical properties of substrate. So the answer to your second question is no.

Your first question is unclear, so I will only generalize by saying that functions tend to be preserved better than sequences. There are relatively few residues in almost any enzyme that have to be absolutely conserved, and that is usually because the reaction chemistry requires particular amino-acid side-chains. There are usually several other residues in or near the active site that are involved in substrate recognition, and they are typically well-conserved but not as much necessarily as catalytic residues. Almost everything else is flexible to a good degree.

Let's say that each human has 5 catalytic residues at the tips of our fingers. If I give you a baseball that has 5 dots on it, almost any person in the world would be able to line up the tips of their fingers to those 5 dots, regardless of how big or small their hands are. This wouldn't be the case if I gave you a ping-pong ball or a basketball, as people with large hands could not handle the former, and those with small hands the latter. The point is that as long as a small number of catalytic residues are preserved and they are spatially organized in a defined way, the exact scaffold (small hands, large hands, young, old, men, women) that brings them to that spatial position is not very important. That's why enzymes can retain the same (or similar) functionality even when their sequences change to a good degree.