About 25 years ago, a promising candidate emerged that could potentially explain poorly understood physiological properties of vertebrate myofibrils—such as their elasticity—which cannot be explained by the acto-myosin based two-filament sliding model. Proteinbiochemical and histological studies identified a protein species with a megadalton size. This protein was named "titin" according to the "titans" of Greek mythology by K. Wang (for reviews, see 1-4). Studies with titin antibodies showed that in situ, the titin epitope map extends over ~2 µm and therefore about the entire half-sarcomeric span. Epitopes within titin’s A-band segment remain attached to the thick filament, whereas epitopes from titin’s I-band segment change their positions with respect to the Z-line. Therefore, in the 1980s, the possibility emerged that muscle contains an unconventional third filament system composed of single titin polypeptides. Workers in the field were excited that this third filament system might have both critical functions for structural integrity for the thick filament and with elastic properties. However, outside the specialized field, this concept was widely mistrusted, since most researchers felt it hard to believe that mammalian cells could handle the expression, folding, and assembly of megadalton-sized protein species.

I entered the titin field at this stage, because I was fascinated by the eventual goal of functionally dissecting this giant molecule using molecular genetic approaches. Our initial cloning of partial titin cDNAs corresponding to ~2% of its full length message identified a single locus in the human genome on chromosome 2 which indeed transcribes a giant mRNA (5). Subsequently, our laboratory performed a tour-de-force cDNA cloning strategy involving a combination of cloning partial cDNAs and their extension. This ultimately identified the 82 kb full-length titin cDNA sequence expressed in the human heart, and an even larger 101 kb splice isoform expressed in human soleus skeletal muscle (6). Our cDNA sequence data predicted that titins are 27,000 to 33,000 residue polypeptides with molecular weights ranging between 2,970 kDa (heart muscle) and 3,700 kDa (soleus skeletal muscle). Therefore, the molecular weight of titin exceeds that of a "normal" proteins by 20- to 100-fold. The cDNA sequence data also revealed beautiful highly regular modular substructures of titin: 299 fibronectin-type and immunoglobulin-like repeats in titin are arranged distinct domain architectures in different regions of the molecule and propose a fine structure for the filament system from which its layout within the sarcomere can be inferred. Titin’s amino terminal, about a 100-kDa region, anchors the protein giant within the Z-disc region of the sarcomere; the following approximately 1,500 kDa region corresponds to its elastic part in the soleus-type titin. Finally, titin’s carboxyterminal 2,000 kDa portion anchors the titin filament within the A-band by multiple interactions with other components of the thick filament. For the I-band portion of titin, the presence of distinct motif families proposed that titin is unlikely to be a homogeneous spring (6). Indeed, as we know now today from single myofibril studies three distinct motifs in titin’s I-band region serve as molecular spring elements that are recruited sequentially by different amounts of stretch at different force levels (7).

The titin sequence also codes for a catalytic serine-threonine kinase domain and phosphorylation motifs, thereby providing hints that the filament may participate in signal transduction (6). During the last two years mounting evidence has been found that titin is indeed not merely a passive spring but is dynamic in structure and function and participates in myofibrillar signaling pathways (for a review, see 8). Since titin is the only molecule long enough to span an entire half-sarcomere, the many titin molecules within a myofibril may sense myofibrillar tension and transmit stretch-dependent information to the myocyte. This flow of information from the myofibril could be involved in the regulation of trophic responses. Mouse models created by modern molecular genetic tools will allow us to study the physiological roles of titin and its myofibrillar ligands in vivo in detail. For example, a knock-out mouse model for the titin associated protein MLP/CsRP3 shows impaired stretch-response, raising the possibility that the titin–MLP complex is required for sensing stretch (9). In a set of separate studies, small deletions within the mouse titin were shown to cause severe myopathies linked to the activation of CARP, a cytokine-like protein which regulates the Nkx2.5 pathway (10, 11). These recent mouse molecular genetic studies have provide early indications that genetic defects in titin and its associated factors perturb signaling within myocytes, and future studies are now required to elucidate titin’s role as a biomechanical sensor in detail.

Siegfried Labeit
Professor for Experimental Medicine
Dept. of Anesthesiology and Intensive Operative Care 
Universitätsklinikum Mannheim 
Mannheim, Germany