
Does the HIV virus infect only T helper cells or can it cross the membrane, and infect, other cell types?

Regardless of its entry route, HIV requires two necessary entry events to successfully infect the cells it encounters. First, one of HIV’s surface proteins must come in contact with and bind to its host receptor, CD4, which is a transmembrane protein expressed only on the outer surface of specific cell types. Second, a coreceptor, either the CCR5 or CXCR4 protein, is needed to provide additional energy that helps the virus fuse successfully with the cell membrane. Note that the quest for additional receptors and entry methods is an active area of biomedical research (1).
Several types of cells fit these two requirements. As suggested in your question, helper T cells are one of these types: They express CD4 and either CCR5 or CXCR4 on their surface. These cells circulate in both the bloodstream and local tissue throughout the body, becoming infected when exposed to HIV virions. Ironically, CD4 helps activate T cells to fight infections, and coreceptors CCR5 and CXCR4 are both chemokine receptors that help accurately guide the body’s immune cells to sites of infected tissue.
Other cells that fit these criteria, and so can also be infected with HIV, are macrophages and dendritic cells. These are also cells of the immune system that frequently survey local tissue and can circulate in the bloodstream. Certain types of macrophages, called microglia, help inspect the central nervous system and can also become infected with HIV.
Importantly, cells that are susceptible to HIV infection are themselves immune cells, making HIV infection particularly dangerous and able to cause AIDS over time. HIV-infected individuals often present to physicians with infectious illnesses that appear only when one is immunocompromised, such as oral infection of Candida albicans, which is not harmful to healthy people. Ultimately, it is various types of acquired illnesses that can become lethal in patients with AIDS. Furthermore, the cells that are susceptible to HIV infection are also difficult to fully eradicate with current highly active antiretroviral therapy (HAART), which is why there is no current cure. For instance, CD4+ T cells can form memory T cells with extremely long life spans that reside in local tissue that drugs cannot reach. When patients are taken off of HAART, these cells can become reactivated and help spread HIV systemically (2). In addition, HIV-infected microglia also persist in the brain despite HAART, causing HIV-associated dementia over time.
Interestingly, humans have evolved mutations in the coreceptor CCR5 that render some people resistant to HIV infection. Specifically, CCR5-Δ32 is a deletion mutation in the coreceptor that can be found in people of Northern European descent (3). It is thought that this mutation evolved a long time ago to help people survive the Black Death and perhaps also smallpox. Recently, biomedical researchers have taken advantage of this natural mutation and successfully provided the closest thing to a cure for HIV infection in a man called the “Berlin patient” (4). He needed a bone marrow transplant for non-HIV–related reasons to reconstitute his immune system, and physicians selected a donor who had the CCR5-Δ32 deletion. After a successful transplant, the Berlin patient has been living without HIV medications, and no virus has been detected in his body, for more than two years.
References
1. Didigu, C. A., and Doms, R. W. 2012. Viruses, 4(2):309–324. PMID: 22470838. http://www.ncbi.nlm.nih.gov/pubmed/224708382
2. Lassen, K., Han, Y., Zhou, Y., Siliciano, J., and Siliciano, R. F. 2004. Trends in Molecular Medicine, 10(11):525–531. PMID: 15519278 http://www.ncbi.nlm.nih.gov/pubmed/15519278
3. Galvani, A. P., and Slatkin, M. 2003. PNAS,100(25):15276–15279. PMID 14645720.
4. HIV Cure Is Closer as Patient's Full Recovery Inspires New Research, http://m.npr.org/news/front/156988650?page=0