The term "enhancer" stems from work by Schaffner and Chambon that demonstrated cis regulatory regions from viruses would "enhance" expression from different basal promoter regions. Later work showed that in a cellular context these elements function both as positive and negative switches (depending on proteins bound, cellular physiology/signaling). A good review from Barolo and Posakony discusses how many signaling switches have this dual nature. A key feature of these elements is their ability to work at a variable distance from the transcriptional start site - in vertebrates, some work on promoters located >1Mbp away. Enhancers can often show promoter specificity; Blackman and colleagues demonstrated that in the 1990s for enhancers of the dpp locus in Drosophila, as have many subsequent studies. Thus we cannot assume that a bit of DNA with activators stuck to it - the basics of an enhancer in an active mode - will affect all neighboring basal promoters. Computational types often prefer to use the general term "cis regulatory element" to denote some bit of sequence that has an influence on transcriptional regulation; these may include boundary element activity, enhancer-promoter looping activity, activity that only works at a close distance to the start site. However, the notion of a distally-acting sequence does actually differentiate what these enhancer sequences are doing biochemically. Incidentally, a number of studies have shown how enhancers can work in trans, from one chromosome to another, thus "cis regulatory" is itself a bit limiting. The frequency with which enhancers work in trans is not yet known; might be rather unusual cases, but we don't have hard numbers on that yet. 

Great question!

Regulatory regions can be promoters, enhancers, silencers, insulators, etc. This is a broad category of "things that regulate transcription".

An enhancer is a specific type of regulatory element that is cis-acting (that is, it acts on the same molecule of DNA where it is located). Enhancers increase the use of specific eukaryotic promoters (they enhance transcription). Enhancers can be located either up or downstream of the promoter region and can be quite far away from the promoter that they work on.

Thank for your replies! A couple more related questions:

1. "The presence of enhancer enhances the expression of pretty much everything around it" (Artur) - so enhancers tend to increase transcription of everything around them, rather than have specificity for a particular promoter?

2. "An enhancer is a specific type of regulatory element that is cis-acting (that is, it acts on the same molecule of DNA where it is located)" (Meredith) - so a regulatory region that increases expression of a gene located on a different chromosome is not an enhancer?? 

Apologies if I'm sound pedantic - I would like to appreciate and subtleties arising from the history of these two phrases.

Thanks again! 

The term "enhancer" stems from work by Schaffner and Chambon that demonstrated cis regulatory regions from viruses would "enhance" expression from different basal promoter regions. Later work showed that in a cellular context these elements function both as positive and negative switches (depending on proteins bound, cellular physiology/signaling). A good review from Barolo and Posakony discusses how many signaling switches have this dual nature. A key feature of these elements is their ability to work at a variable distance from the transcriptional start site - in vertebrates, some work on promoters located >1Mbp away. Enhancers can often show promoter specificity; Blackman and colleagues demonstrated that in the 1990s for enhancers of the dpp locus in Drosophila, as have many subsequent studies. Thus we cannot assume that a bit of DNA with activators stuck to it - the basics of an enhancer in an active mode - will affect all neighboring basal promoters. Computational types often prefer to use the general term "cis regulatory element" to denote some bit of sequence that has an influence on transcriptional regulation; these may include boundary element activity, enhancer-promoter looping activity, activity that only works at a close distance to the start site. However, the notion of a distally-acting sequence does actually differentiate what these enhancer sequences are doing biochemically. Incidentally, a number of studies have shown how enhancers can work in trans, from one chromosome to another, thus "cis regulatory" is itself a bit limiting. The frequency with which enhancers work in trans is not yet known; might be rather unusual cases, but we don't have hard numbers on that yet. 

I agree with David Arnosti.

In response to your question1 above:

Enhancers do have specificity for promoters. In addition to the work of Blackman which David has cited, you could find the evience in Genes Dev. 2001; 15: 2515–2519; Genes Dev. 1996; 10: 1260–1270 and EMBO J. 1994 Jan 15;13(2):400-6.

answer to question 2:

 though cis-acting enhancers are more common, there are few examples of trans-acting enhancer elements. Interactions between promoter of IFN-gamma on chromosome 10 and  regulatory regions on chromosome 11, have been  reported (Nature. 2005 Jun 2;435(7042):637-45)

Also, one trans-acting enhancer element is found to activate an olfactory receptor allele in a olfactory sensory neuron (Cell. 2006; 126: 403–413)

"An enhancer is a specific type of regulatory element that is cis-acting (that is, it acts on the same molecule of DNA where it is located). "

- so, when we know when exactly one of molecule DNA starts and when it ends?

So, if I good understand, one chromosome = one molecule of DNA or not?