The development of cancer invariably involves cellular proteins that have unwanted effects or that become inactive. Therapeutically targeting proteins that are more active than normal is conceptually straightforward — a drug needs to be found that, like a wrench thrown into a machine, halts the malicious activity. By contrast, remedying the loss of a type of protein called a tumour suppressor, which normally opposes cancer growth, presents a conundrum. How can a protein be targeted by drugs if it is no longer present? One way of getting around this problem and finding new drug targets is to identify other effects, such as molecular changes, that occur when a tumour suppressor is lost or does not function. Writing in Nature, Walter et al.1 report their analysis of the role of the tumour-suppressor protein retinoblastoma (Rb), using an experimental system that allowed them to engineer the loss and reactivation of this protein in lung cancer in mice.
Since the discovery of Rb’s tumour-suppressor activity from studies of children with a type of eye cancer, and the subsequent identification of Rb loss or inactivation as a characteristic of many human tumours, this protein has served as a model for studying tumour suppressors in cancer2. Rb can act as a key brake on cell-cycle progression, and it promotes cellular differentiation into specific cell types, which slows the development of cancer. Many cancers block Rb function by increasing the activity of enzymes called cyclin-dependent kinases (Cdks), which inactivate Rb through phosphorylation3 (the addition of phosphate groups to it; Fig. 1a). Anticancer treatments involving inhibitor drugs that target Cdk4 and Cdk6 to prevent them from inactivating Rb have shown some success. However, as would be expected, these inhibitors are not effective if Rb is not expressed in the tumours4.
Walter and colleagues studied the effects of Rb loss in a mouse model of a common form of lung cancer called lung adenocarcinoma. The authors used genetically engineered animals to control whether and when Rb was made in lung cancer cells. This showed that the absence of Rb (a state termed an Rb knockout) promotes cancer development by causing an increase in both tumour growth and its spread to secondary sites in a process called metastasis. The results provide some of the strongest in vivo evidence so far to support a role for Rb loss in driving not only cancer initiation, but also its progression.
The authors observed that the molecular changes that lead to progression and metastasis in this lung cancer model are strikingly different, depending on whether Rb is present or absent. If Rb was missing, compared with the situation in which Rb was present but inactivated by Cdks, the cancer cells had an increased propensity to change their cellular identity to a less-differentiated state that allowed the disease to progress more rapidly. This state was accompanied by the expression of metastasis-promoting proteins such as HMGA2. Such cellular dedifferentiation in the absence of Rb has also been observed in prostate cancer5. Although the precise mechanism that controls cellular plasticity in this scenario is unknown, it might involve the activation of factors involved in cellular reprogramming, such as the transcription-factor protein SOX2, which is normally inhibited by Rb6.
Building on previous studies7,8 that investigated Rb loss in mouse models of cancer, Walter and colleagues studied the effect of reinstating Rb expression in mice with lung cancer months after the Rb-encoding gene had been deleted by genetic engineering. They found that re-expression of Rb promoted the redifferentiation of undifferentiated cancer cells, rendering the tumours less malignant. This observation is important, because it suggests that strategies to reactivate Rb or to manipulate proteins acting downstream of Rb might be therapeutically beneficial for treating cancers that have lost Rb expression.
Another key observation made by Walter and colleagues is that tumours that lack Rb do not upregulate a signalling pathway containing the kinase MAPK, which is often activated in lung adenocarcinoma. This had not been reported previously and is unexpected, given that upregulation of the pathway is a hallmark of this type of cancer. The authors’ data strongly suggest that inactivating Rb is a major objective of MAPK activation in lung cancer that is no longer required in the absence of Rb.
The clinical success of Cdk4 and Cdk6 inhibitors indicates that manipulation of Rb activity can stop cancer progression. However, the development of resistance to such therapies is motivating the search for other drugs that could mimic the reactivation of Rb function, and for new targets in combating cancer cells that have lost Rb expression. Walter and colleagues’ success in identifying molecular and cellular effects of Rb loss, such as cellular dedifferentiation and an absence of upregulation of MAPK signalling, will help with these efforts.
MAPK signalling can inactivate Rb through several mechanisms9. Walter and colleagues’ results indicate that Rb inhibition results from increased MAPK signalling that inactivates10 a Cdk inhibitor protein called p27, thereby activating Cdk2 — a Cdk that can phosphorylate and thus inactivate Rb3. The authors’ observations regarding the interplay between MAPK and Cdk2 implicate Cdk2 as a potential drug target in lung cancers that retain Rb but are resistant to Cdk4 and Cdk6 inhibitors. In this context, the combined inhibition of MAPK and Cdk2, or of Cdk4, Cdk6 and Cdk2, might potently stop the growth of lung tumours by ensuring that Rb remains in its active state and is not associated with high levels of phosphorylation (Fig. 1b).
However, more work is needed to identify further potential drug targets. Efforts should also be made to determine the sensitivity of a person’s cancer to inhibitors of Cdk4 and Cdk6 or Cdk2, including assessing the individual’s genetic profile and monitoring alterations termed epigenetic changes — modifications to DNA or DNA-binding proteins that can affect gene expression. Furthermore, it would be useful to identify targetable proteins or pathways that inhibit the cellular dedifferentiation that occurs after Rb loss (Fig. 1c).
Finally, it will be essential to understand to what extent the principles of lung cancer development and progression are applicable to other types of cancer. The experimental system used by Walter et al. could help to address this challenge, because their results might elucidate the effect of loss of tumour-suppressor proteins in diverse contexts. Studies of Rb have long provided a useful model system in which to investigate fundamental mechanisms of tumour suppression. Now they might also offer a way forward in uncovering new anticancer strategies.
Nature 569, 343-344 (2019)