why the enzyme cant synthesize protein directly from DNA
HOW ABOUT through artificial modification of enzyme to get it
I don't understand what do you want.
Do you mean that the rybozome (and accompanying friends) will be modified to synthesize protein on DNA template?
For what purpose? (one purpose i can think of - DNA is more stable than RNA so would be nice for in vitro assays).
Is this a hypothetical question?
looking for research partners for a novel product?
Or are you just unfamiliar with the biochemical makeup of the living cell?
Evolutionarily speaking, the "RNA world" hypothesis holds that RNA was tthe first of the two molecules to evolve (incidentally, RNA is also chemically more stable - however, all of us who do RNA preps lament because of the ubiquitous nature of RNAses - who *degrade* the RNA). DNA came up later - its double stranded form makes it uniquely suited as a database or repository molecule - damage on one strand can be rectified using the other as template.
In addition, if you look in terms of regulation, having an accessible single strand, easily enzymatically degradeable molecule makes a lot of sense.
In direct response to your question, it might perhaps be possible for a modified ribosomal complex to "read" single strand DNA (you might want to read the ribosome structure papers published some 4 years earlier), but frankly, I don't really see the point of it! :)
The TnT system from Promega allows generation of protein directly from DNA, but still generates an RNA transcript first. The gene of interest is cloned into a special plasmid that contains the proper initiation sequences for the transcription machinery to generate an mRNA which is then immediately transcribed into the corresponding protein.
Your uestion about the microarrays is confusing to me. Microarrays are generally made from master plates of cDNA, so there is a source for generating protein without having to worry about using the 'fixed' molecule on the array itself. Are you wondering about making protein arrays from microarrays of unkowns genes? If so, the problem is that most unknown genes are generated from polyA+ mRNA by priming with T11 primers. This creates a bias in the resulting cDNA towards the three prime end of the transcript, which often does not contain the entire coding sequence of the protein, especially for large transcripts. If you were to try to make protein from EST seuences, you would likely end up with junk, if anything at all.
I am very curious to know more about what you are trying to ask. Please try to rephrase your question more carefully, and I will give it some more consideration.